[
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.95
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 16384,
    "loss": null,
    "initialization": "Fresh deterministic initial backbone, zero reader down projections; no pretrained or validation weights loaded",
    "runtime_optimizations": null
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.95
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 16384,
    "loss": null,
    "initialization": "Parent candidate reader80M full-state; existing Adam slots and all RNG/cursor retained; added context projections zero initialized",
    "runtime_optimizations": null
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.95
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 16384,
    "loss": "Learned normalized mixture of mean final CE and mean first CE; GradNorm controller optimizes stopped-target gradient-norm L1 objective; model uses detached weights",
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.95
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 16384,
    "loss": null,
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [],
    "schedule": {
      "coordinate": "raw_words_exposed_after_update",
      "warmup_words": 1600000,
      "initial_warmup_only": true,
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "dropout_boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "cycle_aligned_stages": [
        {
          "start_words": 0,
          "end_words": 10000000,
          "sequence_length": 128,
          "kind": "constant",
          "start_lr": 0.0008,
          "end_lr": 0.0008
        },
        {
          "start_words": 10000000,
          "end_words": 20000000,
          "sequence_length": 256,
          "kind": "constant",
          "start_lr": 0.0006,
          "end_lr": 0.0006
        },
        {
          "start_words": 20000000,
          "end_words": 30000000,
          "sequence_length": 512,
          "kind": "constant",
          "start_lr": 0.0004,
          "end_lr": 0.0004
        },
        {
          "start_words": 30000000,
          "end_words": 40000000,
          "sequence_length": 128,
          "kind": "constant",
          "start_lr": 0.0007,
          "end_lr": 0.0007
        },
        {
          "start_words": 40000000,
          "end_words": 50000000,
          "sequence_length": 256,
          "kind": "constant",
          "start_lr": 0.0005,
          "end_lr": 0.0005
        },
        {
          "start_words": 50000000,
          "end_words": 60000000,
          "sequence_length": 512,
          "kind": "constant",
          "start_lr": 0.0003,
          "end_lr": 0.0003
        },
        {
          "start_words": 60000000,
          "end_words": 70000000,
          "sequence_length": 128,
          "kind": "constant",
          "start_lr": 0.0005,
          "end_lr": 0.0005
        },
        {
          "start_words": 70000000,
          "end_words": 80000000,
          "sequence_length": 256,
          "kind": "constant",
          "start_lr": 0.0003,
          "end_lr": 0.0003
        },
        {
          "start_words": 80000000,
          "end_words": 90000000,
          "sequence_length": 512,
          "kind": "constant",
          "start_lr": 0.00015,
          "end_lr": 0.00015
        },
        {
          "start_words": 90000000,
          "end_words": 100000000,
          "sequence_length": 512,
          "kind": "cosine",
          "start_lr": 0.00015,
          "end_lr": 6e-05
        }
      ],
      "boundary_semantics": "exact boundary update uses prior stage; first update above boundary uses new stage",
      "rationale": "Preserve proven initial LR and stable second-cycle peak; reduce LR within each cycle and at later cycle peaks; finish 90-100M without restart. Candidate, not established optimum."
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.95
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 16384,
    "loss": null,
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Exact common backbone and knob weights from best Baby4096 beta2 zero-word state; remove reader parameters; rebuild empty Adam with identical coefficients; preserve all RNG and scheduler.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "cosine after 60M only",
      "anneal_after_60m": {
        "start_words": 60000000,
        "end_words": 100000000,
        "start_lr": 0.0006,
        "end_lr": 6e-05
      }
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.95
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 16384,
    "loss": null,
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "cosine after 60M only",
      "anneal_after_60m": {
        "start_words": 60000000,
        "end_words": 100000000,
        "start_lr": 0.0006,
        "end_lr": 6e-05
      }
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.95
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      128,
      128,
      128,
      128,
      128,
      256,
      256,
      512,
      512
    ],
    "batch": 16384,
    "loss": null,
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "cosine after60M",
      "late_cosine": {
        "start_words": 60000000,
        "end_words": 100000000,
        "final_lr": 6e-05
      }
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "embedding_weight_decay": 0.0,
      "embedding_lr_ratio": 1.0
    },
    "sequence": [
      512,
      512,
      512,
      512,
      512,
      512,
      512,
      512,
      512,
      512
    ],
    "batch": 4096,
    "loss": null,
    "initialization": "Original8 zero-word weights and all RNG; empty Adam; attached recipe groups; no pretrained weights",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling",
      "head_chunk": 4096,
      "checkpoint_mean": false,
      "checkpoint_blocks_from_length": 0,
      "compile_hidden": false,
      "one_gpu_job_only": true,
      "checkpoint_teacher": false,
      "compile_tiles": false,
      "emulate_precision_casts": false,
      "hidden_backend": "inductor",
      "checkpoint_ce": false,
      "checkpoint_block_stride": 7
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {}
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "final normalized mixture NLL; no auxiliary loss",
    "initialization": "Q0 zero-word exact common tensors and RNG; only new head matrices initialized independently; empty Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.993608849045455
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 2048,
    "loss": null,
    "initialization": "Exact4096 beta2 parent zero-word weights, empty Adam and RNG; beta2 adjusted for2048 slots",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.95
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": null,
    "initialization": "Exact shared parent0-word full-state: no pretraining; identical initial weights and RNG; empty Adam slots",
    "runtime_optimizations": null
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": null,
    "initialization": "Exact4096 parent zero-word state with empty Adam; only optimizer beta2 replaced",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256
    ],
    "batch": 8192,
    "loss": null,
    "initialization": "Exact4096 beta2 parent zero-word weights, empty Adam, scheduler and all RNG; no optimizer coefficient changes",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Exact best9 zero-word common weights/RNG; append fresh10th block. Empty Adam; no trained weights.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Common7-layer zero-word weights and RNG from best Baby4096 beta2 parent; append one newly initialized eighth block. New block uses existing deterministic name-seeded initializer with cfg.layers8; no pretrained training weights.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Preserve all8-layer zero-word weights and RNG; append one newly initialized ninth block with native9-layer initializer; empty Adam with identical coefficients",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Exact baseline9 zero-word weights, optimizer groups, scheduler, all RNG; history_norm=joint_pair; no pretrained weights",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Exact80M full-state continuation of authorized learned-continuous run; model/Adam/allRNG/data cursor preserved",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "All common weights from the native9 two-global-knob zero-word checkpoint;72 fresh calibrated head gains; empty Adam and preserved parent RNG/cursor",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Fresh0M native7->8->9 lineage; fan-in matrix correction; original9 knob initial values; no pretrained matrices or optimizer. Attention256 output projects to state272; attention variance256 feeds FFN directly. Reader64 down zero.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Fresh0M; same common weights and native7->8->9 fan-scaled lineage as fixed272 Attention256 Reader64. Original9 scalar initial values. Only two additional scalar knobs initialized10000 and0.1; no pretrained weights or optimizer.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.00078,
      0.00076,
      0.00074,
      0.00072,
      0.0007,
      0.00068,
      0.00066,
      0.00064,
      0.00062
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.00078,
        0.00076,
        0.00074,
        0.00072,
        0.0007,
        0.00068,
        0.00066,
        0.00064,
        0.00062
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "stepwise each10M; original1.6M warmup retained",
      "lr_policy": "ten_stage_per10m_right_closed",
      "lr_boundaries": [
        0,
        10000000,
        20000000,
        30000000,
        40000000,
        50000000,
        60000000,
        70000000,
        80000000,
        90000000,
        100000000
      ],
      "dropout_boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ]
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      256,
      256,
      256,
      256,
      256,
      256,
      256,
      256,
      256,
      256
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Fresh0M; same common weights and native7->8->9 fan-scaled lineage as fixed272 Attention256 Reader64. Original9 scalar initial values. Only two additional scalar knobs initialized10000 and0.1; no pretrained weights or optimizer.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Fresh width272 zero-word native7->8->9 lineage; scale linear weights by sqrt(reference_fan_in/current_fan_in); embeddings/vectors/knobs unchanged; 74 fresh knobs at same initial values as width256 learned run",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Fresh width272 zero-word native7->8->9 lineage; scale linear weights by sqrt(reference_fan_in/current_fan_in); embeddings/vectors/knobs unchanged; 2 fresh knobs at same initial values as width256 learned run",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Fresh wider zero-word weights under the same9-layer initialization lineage; original9 RNG and empty Adam/scheduler; no trained weights",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Same zero-word fan-scaled native7->8->9 backbone/Reader and empty Adam/RNG/data/scheduler as preceding no-distance272; overwrite only learned knob raw scalars and per-layer mean gate bias from learned25680M",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "embedding_weight_decay": 0.1,
      "embedding_lr_ratio": 1.0
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": null,
    "initialization": "Same8-layer zero-word full state and RNG; no trained weights; optimizer groups configured per arm",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling",
      "head_chunk": 4096,
      "checkpoint_mean": false,
      "checkpoint_blocks_from_length": 0,
      "compile_hidden": false,
      "one_gpu_job_only": true
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "cosine after60M",
      "late_cosine": {
        "start_words": 60000000,
        "end_words": 100000000,
        "final_lr": 6e-05
      }
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "embedding_weight_decay": 0.0,
      "embedding_lr_ratio": 1.0
    },
    "sequence": [
      512,
      512,
      512,
      512,
      512,
      512,
      512,
      512,
      512,
      512
    ],
    "batch": 4096,
    "loss": null,
    "initialization": "Same8-layer zero-word full state and RNG; no trained weights; optimizer groups configured per arm",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling",
      "head_chunk": 4096,
      "checkpoint_mean": false,
      "checkpoint_blocks_from_length": 0,
      "compile_hidden": false,
      "one_gpu_job_only": true
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "role_overrides": {}
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Native9 zero-word common tensors and Reader32 rows preserved;16 extra Reader up rows seeded;all down weights zero;empty Q0 Adam;original Attention",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {}
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Native9 zero-word common tensors and Reader32 rows preserved;extra Reader rows seeded,all down weights zero;fresh sqrt32 gains and empty per-role Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Native9 zero-word common tensors and Reader32 rows preserved;extra Reader rows seeded,all down weights zero;fresh sqrt32 gains and empty per-role Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.15
        },
        "temperature": {
          "lr_scale": 0.15
        },
        "threshold": {
          "lr_scale": 0.3
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.3
        },
        "ffn_variance": {
          "lr_scale": 0.3
        },
        "attention_gate_bias": {
          "lr_scale": 0.15
        },
        "head_gain": {
          "lr_scale": 0.75
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 0.75,
          "weight_decay": 0.1
        },
        "reader_gate_weight": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.3
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Native9 zero-word common tensors and Reader32 rows preserved;extra Reader rows seeded,all down weights zero;fresh sqrt32 gains and empty per-role Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.6
        },
        "temperature": {
          "lr_scale": 0.6
        },
        "threshold": {
          "lr_scale": 1.0
        },
        "rope": {
          "lr_scale": 0.3
        },
        "candidate_variance": {
          "lr_scale": 1.0
        },
        "ffn_variance": {
          "lr_scale": 1.0
        },
        "attention_gate_bias": {
          "lr_scale": 0.6
        },
        "head_gain": {
          "lr_scale": 1.5
        },
        "reader_up": {
          "lr_scale": 0.75,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.5,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.75,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.75
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Native9 zero-word common tensors and Reader32 rows preserved;extra Reader rows seeded,all down weights zero;fresh sqrt32 gains and empty per-role Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "reader_down": {
          "lr_scale": 0.5,
          "weight_decay": 0.2
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Native9 zero-word common tensors and Reader32 rows preserved;extra Reader rows seeded,all down weights zero;fresh sqrt32 gains and empty per-role Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [],
    "schedule": {},
    "optimizer": {},
    "sequence": null,
    "batch": null,
    "loss": null,
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0008,
      0.0008
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        100000000
      ],
      "lr": [
        0.0008,
        0.0008,
        0.0008
      ],
      "ffn_dropout": [
        0.0,
        0.0,
        0.0
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.95
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      512
    ],
    "batch": 16384,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Fresh named-seed common weights reconstructed from original threshold09 CPU factory; added readers use deterministic up and zero down/gate; no pretrained weights used",
    "runtime_optimizations": null
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Authentic9L MoS3 zero common weights bit-exact; fresh512-to256 compression std1/sqrt512; fresh Adam; parent RNG/data cursor",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Original9 zero-word common tensors and all RNG/data/scheduler retained; 2 global knobs same initial values; 9 received matrices zero; empty Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "selective_qv": {
          "lr_scale": 0.1,
          "weight_decay": 0.0
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Q0 zero-word common tensors exact;new content-selection matrices zero;empty Adam;Q0 zero-word RNG",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [
      {
        "lr": [
          7.000000000000001e-05,
          5.9999999999999995e-05,
          5e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00041999999999999996,
          0.00035999999999999997,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00020999999999999998,
          0.00017999999999999998,
          0.00015
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0007,
          0.0006,
          0.0005
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0007,
          0.0006,
          0.0005
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00046666666666666666,
          0.00039999999999999996,
          0.0003333333333333333
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.00046727272727272725,
          0.0004005194805194805,
          0.0003337662337662338
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.00023408796895213455,
          0.00020064683053040103,
          0.00016720569210866755
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0007,
          0.0006,
          0.0005
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00035,
          0.0003,
          0.00025
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s2855770d582b2b",
    "base_lrs": [
      0.0007,
      0.0006,
      0.0005
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0007,
        0.0006,
        0.0005
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh384 matrix initialization, reference256 fan-in scaling; preserve MoS facet identity halves, native compression scaling and zero mixtures/gates. Copy same-shape initial parameters and training RNG from existing untrained256 MoS2 fixture in RAM; empty Adam. No initial or recovery checkpoint.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Prefix CUDA graph, FP32 TF32off; real-prompt throughput and parity checked before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent Q3+MoS2 zero-word full state common weights/RNG; nine new received-gate matrices zero; empty Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh zero-word run: preserve all original9L zero weights exactly; independently initialize block9 with existing per-name seeds and native10L residual std; no trained weights. Parent RNG/cursor retained and fresh empty Adam for10L.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "OriginalMoS3 zero common weights exact; fresh2304-to256 compression std1/sqrt2304; layer logits0; dynamic query0,keyN(0,1/sqrt256); freshAdam,parentRNG/data cursor",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s34bda1cc55404f",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh320 native fan-in branch compression; C3 copies C2 into h8/h9 columns and zero h7 columns, then all columns learn. Preserve facet h9 identity; width-scale context half. Existing untrained256 zero fixture only for unchanged-shape parameters/RNG in RAM. Fresh Adam; no new initial/recovery weight files.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight",
          "multi_output.mod2.weight",
          "multi_output.mod3.weight"
        ],
        "label": "embed.weight, multi_output.mod2.weight, multi_output.mod3.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.base.weight",
          "multi_output.delta9.weight",
          "multi_output.delta8.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.base.weight, multi_output.delta9.weight, multi_output.delta8.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s13f50438bfcaa8",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh0M; same parent320 backbone and facet initialization. Shared h9 projection uses parent compress1 initial draw; separate64-wide signed delta encoders native fan-in; scale/shift matrices start at zero and learn. Existing zero256 fixture read in RAM only for unchanged shapes and training RNG. No initial/recovery files.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh0M as separate987 parent, same parameter draws and width scaling; replace output input tensors only. Existing0M fixtures may be read in RAM; no new initial/recovery weights.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0005333333333333334,
          0.00046666666666666666,
          0.00039999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0005340259740259741,
          0.00046727272727272725,
          0.0004005194805194805
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.00026752910737386806,
          0.00023408796895213455,
          0.00020064683053040103
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s3aeb25fa874167",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "New0M native per-name seed, explicit width initialization, same-shape parent zero parameters, parent initial training RNG; fresh Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Validate last1 exact recurrence and batched prefix graph before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0004007797270955166,
          0.000350682261208577,
          0.00030058479532163744
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0002009718172983479,
          0.0001758503401360544,
          0.0001507288629737609
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:scb26ff4b7929f9",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh per-name native initialization, inherited reference256/FFN896 matrix fan-in scaling. Read existing zero-word fixture in RAM only for unchanged shapes and initial training RNG. Preserve last1 facet RNG convention. Do not load or rescale trained384 weights. Do not save an initial checkpoint.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Pending512 GPU runtime acceptance"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight",
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "blocks.8.ffn.spread_input.weight",
          "lm_head.weight",
          "multi_output.compress.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "embed.weight, blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.spread_input.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s32ea4c5c131558",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role_lr_scale unchanged; width affects initialization only"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "New0M native per-name seed, explicit width initialization, same-shape parent zero parameters, parent initial training RNG; fresh Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "embed.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0007529411764705883,
          0.0006588235294117646,
          0.0005647058823529411
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.ffn.base.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.ffn.base.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0007466666666666667,
          0.0006533333333333333,
          0.00056
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.down.weight",
          "blocks.1.ffn.down.weight",
          "blocks.2.ffn.down.weight",
          "blocks.3.ffn.down.weight",
          "blocks.4.ffn.down.weight",
          "blocks.5.ffn.down.weight",
          "blocks.6.ffn.down.weight",
          "blocks.7.ffn.down.weight",
          "blocks.8.ffn.down.weight"
        ],
        "label": "blocks.*.ffn.down.weight"
      },
      {
        "lr": [
          0.0003766848816029144,
          0.0003295992714025501,
          0.00028251366120218575
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18841f8f6f87b",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "role_lr_scale * per_matrix_reference_fan_in / current_fan_in"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "New0M native per-name seed, explicit width initialization, same-shape parent zero parameters, parent initial training RNG; fresh Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s2a2d7fc8bd8b74",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh320 matrices with reference256 fan-in scaling; preserve facet identity halves, native compression initialization and zero gates/mixture. Read existing untrained256 last2 parameters/RNG in RAM; fresh Adam. No new initial/recovery weights.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0005333333333333334,
          0.00046666666666666666,
          0.00039999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0005340259740259741,
          0.00046727272727272725,
          0.0004005194805194805
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.00026752910737386806,
          0.00023408796895213455,
          0.00020064683053040103
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb7f38614a19421",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "New0M native per-name seed, explicit width initialization, same-shape parent zero parameters, parent initial training RNG; fresh Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Authentic9L MoS3 zero common weights bit-exact; fresh1024-to256 compression std1/sqrt1024; fresh Adam; preserve parent RNG/data cursor",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Authentic9L MoS3 zero common weights bit-exact; fresh1280-to256 compression std1/sqrt1280; fresh Adam; parent RNG/data cursor",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.facet1.weight",
          "multi_output.compress2.weight",
          "multi_output.facet2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.facet1.weight, multi_output.compress2.weight, multi_output.facet2.weight, multi_output.compress3.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:se8f2bac984b3c4",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh0M. Parent320 backbone, vocabulary and all three facet initial weights preserved. compress1 retains parent C1; separated C2/C3 use independent left/right320-column blocks of parent random C2, multiplied by sqrt(2) for single-layer fan-in. Each facet sees its own layer plus its own nonlinear context. Mixing starts uniform; no common h9 append. Existing untrained256 fixture only for unchanged-shape knobs and RNG in RAM.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_gate_encoder": {
          "lr_scale": 1.0,
          "weight_decay": 0.1
        },
        "ffn_gate_output": {
          "lr_scale": 1.0,
          "weight_decay": 0.0
        },
        "ffn_gate_bias": {
          "lr_scale": 1.0,
          "weight_decay": 0.0
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Authentic zero-word MoS3 common weights and RNG; fresh unit-start FFN input/output gates; empty Adam.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Authentic received-gate+MoS2 zero-word common weights; mixing zero row appended; independent third facet. Fresh optimizer required, parent RNG preserved.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "0.5*(draft NLL + proposal NLL) + 0.5*(predicted_gain - stop_gradient(draft NLL - proposal NLL))^2; judge inputs detached",
    "initialization": "Parent MoS3 authentic zero-word common weights loaded read-only into RAM; new review/judge freshly initialized; fresh optimizer",
    "runtime_optimizations": null
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0005333333333333334,
          0.00046666666666666666,
          0.00039999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.facet4.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.facet4.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0005340259740259741,
          0.00046727272727272725,
          0.0004005194805194805
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.00026752910737386806,
          0.00023408796895213455,
          0.00020064683053040103
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s2108c431bc9f27",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh384 native+fan-in initialization; unchanged-shape parameters and training RNG read only from existing zero-word256 MoS4 fixture. Four identity main halves; independent random context halves. Empty Adam. No new initial/recovery saves.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Prefix CUDA graph, FP32 TF32off; real-prompt throughput and parity checked before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Authentic zero-word MoS3 common weights and RNG; independent facet4; mixing adds one zero row; empty Adam. Fresh0M training.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0005333333333333334,
          0.00046666666666666666,
          0.00039999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.facet1.weight",
          "multi_output.compress2.weight",
          "multi_output.facet2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet3.weight",
          "multi_output.compress4.weight",
          "multi_output.facet4.weight",
          "multi_output.compress5.weight",
          "multi_output.facet5.weight",
          "multi_output.compress6.weight",
          "multi_output.facet6.weight",
          "multi_output.compress7.weight",
          "multi_output.facet7.weight",
          "multi_output.compress8.weight",
          "multi_output.facet8.weight",
          "multi_output.compress9.weight",
          "multi_output.facet9.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.facet1.weight, multi_output.compress2.weight, multi_output.facet2.weight, multi_output.compress3.weight, multi_output.facet3.weight, multi_output.compress4.weight, multi_output.facet4.weight, multi_output.compress5.weight, multi_output.facet5.weight, multi_output.compress6.weight, multi_output.facet6.weight, multi_output.compress7.weight, multi_output.facet7.weight, multi_output.compress8.weight, multi_output.facet8.weight, multi_output.compress9.weight, multi_output.facet9.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0005340259740259741,
          0.00046727272727272725,
          0.0004005194805194805
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.00026752910737386806,
          0.00023408796895213455,
          0.00020064683053040103
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sad126810429860",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh384. Each branch reads exactly one layer h1..h9; independent d-to-d fan-in compression and GELU/sphere, own layer state plus own compressed state to facet. No shared h9 append. Shared vocabulary projection; learned per-token9-way probability mixture starts equal. Existing untrained256 fixture only for unchanged-shape knobs/RNG in RAM. No initial/recovery checkpoint.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Q3 zero-word full-state common weights and RNG; dedicated generator initializes new output; empty Adam; no trained checkpoint continuation",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_gate_encoder": {
          "lr_scale": 1.0,
          "weight_decay": 0.1
        },
        "ffn_gate_output": {
          "lr_scale": 1.0,
          "weight_decay": 0.0
        },
        "ffn_gate_bias": {
          "lr_scale": 1.0,
          "weight_decay": 0.0
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Single linear output cross entropy on final pass only; no auxiliary loss",
    "initialization": "Authentic parent0M common weights and RNG, new FFN gates gain1, fresh empty Adam; validation updates discarded",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Single linear output cross entropy on final pass only; no auxiliary loss",
    "initialization": "Parent received-gate zero-word common weights and RNG exactly; discard only multi_output.*; empty Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "cosine after 60M only",
      "anneal_after_60m": {
        "start_words": 60000000,
        "end_words": 100000000,
        "start_lr": 0.0006,
        "end_lr": 0.0001
      }
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.95
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 16384,
    "loss": null,
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "All common tensors copied bit-exact from original9 zero-word full-state; only2 fresh scalar parameters; empty Adam; all parent RNG and cursor retained",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Native9 two-global-knob zero-word common weights, preserved RNG/cursor;72 head gains initialized sqrt(32);empty four-group AdamW",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "cosine after60M",
      "late_cosine": {
        "start_words": 60000000,
        "end_words": 100000000,
        "final_lr": 6e-05
      }
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "embedding_weight_decay": 0.0,
      "embedding_lr_ratio": 1.0
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": null,
    "initialization": "Use the completed8-layer baseline zero-word full state; preserve all initial weights and RNG; apply repaired mechanism and embedding optimizer grouping",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final next-token CE only; gradients through all three core repetitions and anchor; no detach",
    "initialization": "Fresh name-seeded native VISTA initialization; residual out/down scaled for effective depth16; adapter initialized[.9I,.1I]. No pretrained weights; RNG seed20260907.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling",
      "head_chunk": 4096,
      "checkpoint_blocks": false,
      "compile_hidden": false,
      "single_gpu": true,
      "target_head": "only valid labels; host-precomputed indices",
      "speed_preflight": "real128/256/512 batches; padded reference,unpadded eager,cudagraph hidden; strict all-parameter gradients and exact dropout resume"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight",
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "norm.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight"
        ],
        "label": "embed.weight, blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": null,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicit200namedLRs:unchanged384 major x0.9,other x1.05;numericWDsame;MoS includedmajor"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged",
      "main_matrix_lr_schedule": {
        "values": [
          0.0007,
          0.0006,
          0.0005
        ],
        "boundaries": [
          30000000,
          60000000
        ],
        "selection": "role matrices except embed.weight; Candidate Reader and scalar/bias/Norm parameters excluded",
        "semantics": "absolute effective LR after warmup; applied before optimizer.step"
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh320 native fan-in branch compression; C3 copies C2 into h8/h9 columns and zero h7 columns, then all columns learn. Preserve facet h9 identity; width-scale context half. Existing untrained256 zero fixture only for unchanged-shape parameters/RNG in RAM. Fresh Adam; no new initial/recovery weight files.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged",
      "selector": {
        "base_lr": 0.0006,
        "matrix_lr_scale": 0.8,
        "bias_lr_scale": 1.0,
        "matrix_weight_decay": 0.1,
        "vector_weight_decay": 0.0,
        "warmup_words": 1600000,
        "clip": 1.0
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "equal mean of all three component next-token NLLs to body; detached segment preference loss to selector",
    "initialization": "Fresh320 native fan-in branch compression; C3 copies C2 into h8/h9 columns and zero h7 columns, then all columns learn. Preserve facet h9 identity; width-scale context half. Existing untrained256 zero fixture only for unchanged-shape parameters/RNG in RAM. Fresh Adam; no new initial/recovery weight files.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch",
      "selector": "reuse vocabulary projection; features at segment starts only",
      "selected_loss_only": false
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight",
          "multi_output.router.up.bias",
          "multi_output.router.down.bias"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, multi_output.router.up.bias, multi_output.router.down.bias"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.router.up.weight",
          "multi_output.router.down.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.router.up.weight, multi_output.router.down.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s7c3b6bb8be15ca",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged",
      "router_gradient_clip": 1.0,
      "direct_router": {
        "direct_router_matrix": {
          "lr_scale": 0.8,
          "weight_decay": 0.1
        },
        "direct_router_bias": {
          "lr_scale": 1.0,
          "weight_decay": 0.0
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Body: final mixed NLL with detached router weights; Router: CE from detached softmax of three gold-token log probabilities; disjoint parameter gradients and clipping; equal valid-token denominators.",
    "initialization": "Fresh320 native fan-in branch compression; C3 copies C2 into h8/h9 columns and zero h7 columns, then all columns learn. Preserve facet h9 identity; width-scale context half. Existing untrained256 zero fixture only for unchanged-shape parameters/RNG in RAM. Fresh Adam; no new initial/recovery weight files.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Reconstruct original parent320 zero-word tensors in RAM, copy all existing coordinates exactly; preserve g/u row halves; added input rows use parent std, added down column zero. Restore original training RNG. No 0M checkpoint written.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "Frozen parent320 numeric fan-in denominators; every common parameter keeps parent effective LR and WD. New bypass baseLR*0.3 WD0."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std and LR coefficients, per-name seed and same-shaped tensors, parent zero-word training RNG; resized FFN tensors freshly generated; new signed bypass=0. No post-training rescale.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Inherited parent320 numeric coefficients; FFNdown uses896/1120=0.8 even though actual fan-in is1280; other matrices retain parent rules"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 seed and per-name initialization; same per-weight std and numeric width multipliers, FFNdown anchored at1120; changed43 shapes initialized fresh. Unchanged tensors checked bitwise. Restore the same zero-word training RNG fixture in RAM.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Same nonlegacy seed-reset policy as seed731946281; model initialization and training/dropout RNG use the new seed; frozen data order unchanged.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Same nonlegacy seed-reset policy as seed731946281; model initialization and training/dropout RNG use the new seed; frozen data order unchanged.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Model tensors follow matchedseed parent; seed20260907 restores original parent legacy0M RNG in RAM; seed731946281 uses manualseed reset exactly as its parent. No new initial checkpoint.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        },
        "ffn_context": {
          "lr_scale": 0.3,
          "weight_decay": 0.1
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Same-seed parent tensors and training RNG; A and b zero; no additional RNG consumed; no initial checkpoint.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Model tensors follow matchedseed parent; seed20260907 restores original parent legacy0M RNG in RAM; seed731946281 uses manualseed reset exactly as its parent. No new initial checkpoint.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [],
    "schedule": {
      "coordinate": "gate_words=words_after-80000000",
      "warmup_words": 1600000,
      "base_lr": 0.0006,
      "decay": "none",
      "ffn_dropout": 0.0
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "fused": true,
      "gradient_clip": 1.0,
      "matrix_lr_scale": 0.8,
      "bias_lr_scale": 1.0,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "parent_lr": 0,
      "parent_weight_decay": 0
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "final mixed next-token NLL only; gate gradients only; no label-derived auxiliary teacher",
    "initialization": "parent80M weights frozen; fresh readout gate and fresh Adam; deterministic seed; remaining epoch8/9 files",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [],
    "schedule": {
      "coordinate": "gate_words=words_after-80000000",
      "warmup_words": 1600000,
      "base_lr": 0.0006,
      "decay": "none",
      "ffn_dropout": 0.0
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "fused": true,
      "gradient_clip": 1.0,
      "matrix_lr_scale": 0.8,
      "bias_lr_scale": 1.0,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "parent_lr": 0,
      "parent_weight_decay": 0
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "token-count and NLL-gap weighted all-pairs logistic preference; normalized by total gap mass",
    "initialization": "parent80M weights frozen; fresh readout gate and fresh Adam; deterministic seed; remaining epoch8/9 files",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch",
      "selector": "reuse vocabulary logits; statistics and selector MLP only at segment starts",
      "rejected": "mean-tile compilation changed 3-update selector weights; excluded"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight",
          "multi_output.prediction_gate.up.bias",
          "multi_output.prediction_gate.down.bias"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, multi_output.prediction_gate.up.bias, multi_output.prediction_gate.down.bias"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight",
          "multi_output.prediction_gate.up.weight",
          "multi_output.prediction_gate.down.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight, multi_output.prediction_gate.up.weight, multi_output.prediction_gate.down.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s06e123a8f27fe5",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Original200 named LR/WD unchanged; newgate4 parameters LR multipliedf and WD dividedf preserving effective decay",
      "effective_parameter_policy": {
        "mode": "frozen_named_parent",
        "profile": "references/selected_OPTIMIZER_PARAMETER_RULES.json",
        "sha256": "4b196aaec1cb170e76baaca836c50b491f14e2e86ba1b961e8215c7018fa28f8",
        "parent_contract": "references/parent320_CONTRACT.json",
        "parent_contract_sha256": "865064dc227a1e089d5ccf170b3b94256a6c42b0bb57525c367a38811d582b46",
        "target_geometry": {
          "width": 320,
          "hidden": 1120,
          "layers": 9,
          "q_heads": 10,
          "head_dim": 32,
          "reader_width": 48,
          "mixture_gate_hidden": 320
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "All200 shared parameters exactly initialized as parent320; new gate up native fan-in normal, all biases zero, down matrix zero. Gate allocation uses isolated RNG. No pretrained checkpoint.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "All parent204 named coefficients retained, then only newgate four LR multiplied f and WD divided f;200 earlier parameters exact",
      "effective_parameter_policy": {
        "mode": "frozen_named_parent",
        "profile": "references/extended320_OPTIMIZER_PARAMETER_RULES.json",
        "sha256": "61ed496a87bfbea9b48d8de7caec57ef3b69af7911538c54ff0ae8e645edee91",
        "parent_contract": "references/parent320_CONTRACT.json",
        "parent_contract_sha256": "865064dc227a1e089d5ccf170b3b94256a6c42b0bb57525c367a38811d582b46",
        "target_geometry": {
          "width": 320,
          "hidden": 1120,
          "layers": 9,
          "q_heads": 10,
          "head_dim": 32,
          "reader_width": 48,
          "mixture_gate_hidden": 320
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "All200 shared parameters exactly initialized as parent320; new gate up native fan-in normal, all biases zero, down matrix zero. Gate allocation uses isolated RNG. No pretrained checkpoint.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged",
      "readout_gate": {
        "base_lr": 0.0006,
        "matrix_lr_scale": 0.8,
        "bias_lr_scale": 1.0,
        "matrix_weight_decay": 0.1,
        "vector_weight_decay": 0.0,
        "warmup_words": 1600000,
        "clip": 1.0
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "final mixed next-token NLL only; same loss updates body and readout gate",
    "initialization": "Fresh320 native fan-in branch compression; C3 copies C2 into h8/h9 columns and zero h7 columns, then all columns learn. Preserve facet h9 identity; width-scale context half. Existing untrained256 zero fixture only for unchanged-shape parameters/RNG in RAM. Fresh Adam; no new initial/recovery weight files.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch",
      "gate": "one shared vocabulary GEMM; detached statistics per valid prediction; optimize after predecessor workflow complete"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 exact0M and parent training RNG; concatenate3 identical copies of parent0M lm_head.weight; independent coefficients thereafter; no new initialization checkpoint",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Identical parent per-parameter initialization formula and zero/identity constants; seed731946281. Parent legacy fixture copies only identical constants. No old RNG restoration; reset global Python/NumPy/Torch CPU/CUDA RNG to alternate seed before first update.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.0.ffn.bypass",
          "blocks.1.attn_gate.proj.bias",
          "blocks.1.ffn.bypass",
          "blocks.2.attn_gate.proj.bias",
          "blocks.2.ffn.bypass",
          "blocks.3.attn_gate.proj.bias",
          "blocks.3.ffn.bypass",
          "blocks.4.attn_gate.proj.bias",
          "blocks.4.ffn.bypass",
          "blocks.5.attn_gate.proj.bias",
          "blocks.5.ffn.bypass",
          "blocks.6.attn_gate.proj.bias",
          "blocks.6.ffn.bypass",
          "blocks.7.attn_gate.proj.bias",
          "blocks.7.ffn.bypass",
          "blocks.8.attn_gate.proj.bias",
          "blocks.8.ffn.bypass"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias, blocks.*.ffn.bypass"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "Exact parent320 per-parameter lr_scale and weight_decay table; current-width initialization remains fan-in based and is independent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent336 factory and fan-in initialization formula retained; FFN hidden1176to1120; down init factor sqrt(896/1120); all effective parent LR/WD fixed. Changed shapes alter RNG draws, not claimed identical initial tensors.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.0.ffn.bypass",
          "blocks.1.attn_gate.proj.bias",
          "blocks.1.ffn.bypass",
          "blocks.2.attn_gate.proj.bias",
          "blocks.2.ffn.bypass",
          "blocks.3.attn_gate.proj.bias",
          "blocks.3.ffn.bypass",
          "blocks.4.attn_gate.proj.bias",
          "blocks.4.ffn.bypass",
          "blocks.5.attn_gate.proj.bias",
          "blocks.5.ffn.bypass",
          "blocks.6.attn_gate.proj.bias",
          "blocks.6.ffn.bypass",
          "blocks.7.attn_gate.proj.bias",
          "blocks.7.ffn.bypass",
          "blocks.8.attn_gate.proj.bias",
          "blocks.8.ffn.bypass"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias, blocks.*.ffn.bypass"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "Exact parent320 per-parameter lr_scale and weight_decay table; current-width initialization remains fan-in based and is independent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Same parent initialization formula at width336/hidden1176; changing tensor shapes changes random draws. Same seed and paired-parent training RNG reset policy; no initial checkpoint.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.0.ffn.bypass",
          "blocks.1.attn_gate.proj.bias",
          "blocks.1.ffn.bypass",
          "blocks.2.attn_gate.proj.bias",
          "blocks.2.ffn.bypass",
          "blocks.3.attn_gate.proj.bias",
          "blocks.3.ffn.bypass",
          "blocks.4.attn_gate.proj.bias",
          "blocks.4.ffn.bypass",
          "blocks.5.attn_gate.proj.bias",
          "blocks.5.ffn.bypass",
          "blocks.6.attn_gate.proj.bias",
          "blocks.6.ffn.bypass",
          "blocks.7.attn_gate.proj.bias",
          "blocks.7.ffn.bypass",
          "blocks.8.attn_gate.proj.bias",
          "blocks.8.ffn.bypass"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias, blocks.*.ffn.bypass"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "Exact parent320 per-parameter lr_scale and weight_decay table; current-width initialization remains fan-in based and is independent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent336 factory and fan-in initialization formula retained; FFN hidden1176to1232; down init factor sqrt(896/1232); all effective parent LR/WD fixed. Changed shapes alter RNG draws, not claimed identical initial tensors.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.0.ffn.bypass",
          "blocks.1.attn_gate.proj.bias",
          "blocks.1.ffn.bypass",
          "blocks.2.attn_gate.proj.bias",
          "blocks.2.ffn.bypass",
          "blocks.3.attn_gate.proj.bias",
          "blocks.3.ffn.bypass",
          "blocks.4.attn_gate.proj.bias",
          "blocks.4.ffn.bypass",
          "blocks.5.attn_gate.proj.bias",
          "blocks.5.ffn.bypass",
          "blocks.6.attn_gate.proj.bias",
          "blocks.6.ffn.bypass",
          "blocks.7.attn_gate.proj.bias",
          "blocks.7.ffn.bypass",
          "blocks.8.attn_gate.proj.bias",
          "blocks.8.ffn.bypass"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias, blocks.*.ffn.bypass"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "Exact parent320 per-parameter lr_scale and weight_decay table; current-width initialization remains fan-in based and is independent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent336/1232 fan-in initialization formula retained at FFN1288; down init scale sqrt(896/1232) to sqrt(896/1288); shape-dependent RNG draws change. All effective parent LR/WD unchanged.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "Exact parent320 per-parameter lr_scale and weight_decay table; current-width initialization remains fan-in based and is independent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "One causal pass, final MoS3 probability NLL; same label mask and denominator; no auxiliary loss.",
    "initialization": "Parent336 same common seeded tensors and fan-in initialization; remove Candidate Reader, prediction-feedback scalars, FFN bypass; restore same historical training RNG before first update.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Single-pass causal FP32 with chunked vocabulary and original left-to-right score sum; GPU prefix verification before training."
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Exact parent320 per-parameter lr_scale and weight_decay table; current-width initialization remains fan-in based and is independent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Same parent336 FFN1232 initialization formula and common tensors; remove zero-initialized FFN bypass only. Restore same legacy0M training RNG; no weights loaded.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.0.ffn.bypass",
          "blocks.1.attn_gate.proj.bias",
          "blocks.1.ffn.bypass",
          "blocks.2.attn_gate.proj.bias",
          "blocks.2.ffn.bypass",
          "blocks.3.attn_gate.proj.bias",
          "blocks.3.ffn.bypass",
          "blocks.4.attn_gate.proj.bias",
          "blocks.4.ffn.bypass",
          "blocks.5.attn_gate.proj.bias",
          "blocks.5.ffn.bypass",
          "blocks.6.attn_gate.proj.bias",
          "blocks.6.ffn.bypass",
          "blocks.7.attn_gate.proj.bias",
          "blocks.7.ffn.bypass",
          "blocks.8.attn_gate.proj.bias",
          "blocks.8.ffn.bypass"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias, blocks.*.ffn.bypass"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "Exact parent320 per-parameter lr_scale and weight_decay table; current-width initialization remains fan-in based and is independent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Same fan-in formula as parent336 at width352/hidden1232. Numeric init factors and shape-dependent RNG draws change; exact parent320 effective LR/WD table retained.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005066666666666667,
          0.0004433333333333333,
          0.00037999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "A major95-other100; only embedding_minus5 x0.95; remaining named LRs and numeric WD unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005066666666666667,
          0.0004433333333333333,
          0.00037999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "A major95-other100; only reader_transform_plus5 x1.05; remaining named LRs and numeric WD unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005066666666666667,
          0.0004433333333333333,
          0.00037999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006084735202492212,
          0.0005324143302180685,
          0.0004563551401869158
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "A major95-other100; only variance_input_minus5 x0.95; remaining named LRs and numeric WD unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, norm.weight"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight"
        ],
        "label": "blocks.*.an.weight, blocks.*.fn.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s899a9612ed2762",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Parent#3 all200 explicit effectiveLRs; specified8 roles48 tensors restored to#31 LR; other152 tensor LR andall200WD unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight"
        ],
        "label": "embed.weight, blocks.*.an.weight, blocks.*.fn.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": null,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": null,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain",
          "norm.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.attention_log_gain, norm.weight, blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "#269 + FFN spread_input 9 weights at baseline +5%; all other named LR/WD unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, norm.weight"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight"
        ],
        "label": "blocks.*.an.weight, blocks.*.fn.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s1b1ded5333b79a",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "#268 all200explicitLRs; alpha,tau,Attentiongatebias11tensorLRrestoredto#31baseline+5%;other189LR/all200WDunchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": null,
        "names": [
          "lm_head.weight",
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "lm_head.weight, feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": null,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "B198namedLRsunchanged;inputembedding0.00082;vocabularyprojection0.000504;numericWDunchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": null,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight",
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "blocks.8.candidate_reader.down.weight",
          "norm.weight"
        ],
        "label": "embed.weight, blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, blocks.*.candidate_reader.down.weight, norm.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "198 approved explicit FINAL coefficients applied once after original role/width resolution; width initialization stays original highest1344last2"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "New0M native per-name seed, explicit width initialization, same-shape parent zero parameters, parent initial training RNG; fresh Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "embed.weight, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.1.qk.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.2.qk.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.3.qk.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.4.qk.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.5.qk.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.6.qk.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.7.qk.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.8.qk.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "multi_output.mixing.weight",
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "blocks.*.qk.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, multi_output.mixing.weight, feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.0005,
          0.0004375,
          0.00037499999999999995
        ],
        "wd": null,
        "names": [
          "blocks.0.v.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.v.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.2.v.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.3.v.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.4.v.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.5.v.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.6.v.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.7.v.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.8.v.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "lm_head.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "candidate_variance_raw"
        ],
        "label": "blocks.*.v.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, lm_head.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, candidate_variance_raw"
      },
      {
        "lr": [
          0.00045599999999999997,
          0.00039899999999999994,
          0.00034199999999999996
        ],
        "wd": null,
        "names": [
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight"
        ],
        "label": "multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight"
      },
      {
        "lr": [
          0.0006084735202492212,
          0.0005324143302180685,
          0.0004563551401869158
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.00052,
          0.00045499999999999995,
          0.00038999999999999994
        ],
        "wd": null,
        "names": [
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw"
        ],
        "label": "feedback_beta, temperature_raw"
      },
      {
        "lr": [
          0.00022799999999999999,
          0.00019949999999999997,
          0.00017099999999999998
        ],
        "wd": null,
        "names": [
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": null,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain"
        ],
        "label": "blocks.*.attention_log_gain"
      },
      {
        "lr": [
          0.00034,
          0.00029749999999999997,
          0.00025499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "49 approved final per-name LR repairs;151 final LR scales inherited; override applied after role/width factors exactly once"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005333333333333334,
          0.00046666666666666666,
          0.00039999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Parent384 FFN1408 numeric per-name coefficients retained, except explicit103 major-matrix LR scales0.8->2/3; WD untouched"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005333333333333334,
          0.00046666666666666666,
          0.00039999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0005066666666666667,
          0.0004433333333333333,
          0.00037999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight"
        ],
        "label": "blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight"
        ],
        "label": "blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Retain fineV2 entire recipe;only9FFNdown LRs match corresponding inputLR;original384 baseline has34modifiedtensors total;allnumericWDfixed"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005333333333333334,
          0.00046666666666666666,
          0.00039999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0005066666666666667,
          0.0004433333333333333,
          0.00037999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.base.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight"
        ],
        "label": "blocks.*.ffn.base.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight"
        ],
        "label": "blocks.*.ffn.base.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Original384/1408 numeric coefficients preserved; only25 FFNinputtensors x0.90 or0.95 bylayer;Router restored toparent; allnumericWD unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005,
          0.0004375,
          0.00037499999999999995
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight"
      },
      {
        "lr": [
          0.00062,
          0.0005424999999999999,
          0.0004649999999999999
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.00052,
          0.00045499999999999995,
          0.00038999999999999994
        ],
        "wd": 0.1,
        "names": [
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight"
      },
      {
        "lr": [
          0.00045000000000000004,
          0.00039375,
          0.00033749999999999996
        ],
        "wd": 0.1,
        "names": [
          "multi_output.mixing.weight"
        ],
        "label": "multi_output.mixing.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0005,
          0.0004375,
          0.00037499999999999995
        ],
        "wd": 0.0,
        "names": [
          "candidate_variance_raw"
        ],
        "label": "candidate_variance_raw"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain"
        ],
        "label": "blocks.*.attention_log_gain"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00034,
          0.00029749999999999997,
          0.00025499999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "All200 final actualLR overrides from approved global recipe; applied after role and width factors, never width-scaled twice"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.2.qk.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.3.qk.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.4.qk.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.5.qk.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.6.qk.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.7.qk.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.8.qk.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.v.weight",
          "blocks.0.out.weight",
          "blocks.1.v.weight",
          "blocks.1.out.weight",
          "blocks.2.v.weight",
          "blocks.2.out.weight",
          "blocks.3.v.weight",
          "blocks.3.out.weight",
          "blocks.4.v.weight",
          "blocks.4.out.weight",
          "blocks.5.v.weight",
          "blocks.5.out.weight",
          "blocks.6.v.weight",
          "blocks.6.out.weight",
          "blocks.7.v.weight",
          "blocks.7.out.weight",
          "blocks.8.v.weight",
          "blocks.8.out.weight"
        ],
        "label": "blocks.*.v.weight, blocks.*.out.weight"
      },
      {
        "lr": [
          0.0006084735202492212,
          0.0005324143302180685,
          0.0004563551401869158
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.000552,
          0.000483,
          0.0004139999999999999
        ],
        "wd": 0.1,
        "names": [
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00022799999999999999,
          0.00019949999999999997,
          0.00017099999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "temperature_raw"
        ],
        "label": "temperature_raw"
      },
      {
        "lr": [
          0.000528,
          0.000462,
          0.000396
        ],
        "wd": 0.0,
        "names": [
          "candidate_variance_raw"
        ],
        "label": "candidate_variance_raw"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain"
        ],
        "label": "blocks.*.attention_log_gain"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight"
        ],
        "label": "blocks.*.an.weight, blocks.*.fn.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00078,
          0.0006825,
          0.0005849999999999999
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00044000000000000007,
          0.00038500000000000003,
          0.00033
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "norm.weight"
        ],
        "label": "norm.weight"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicit named LR: V/out504; DeltaRouter552; CandidateDown780; CandidateGate440; all other139 tensor LRs and all numeric WD inherited from Router125 parent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006084735202492212,
          0.0005324143302180685,
          0.0004563551401869158
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00022799999999999999,
          0.00019949999999999997,
          0.00017099999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "temperature_raw"
        ],
        "label": "temperature_raw"
      },
      {
        "lr": [
          0.000528,
          0.000462,
          0.000396
        ],
        "wd": 0.0,
        "names": [
          "candidate_variance_raw"
        ],
        "label": "candidate_variance_raw"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain"
        ],
        "label": "blocks.*.attention_log_gain"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight"
        ],
        "label": "blocks.*.an.weight, blocks.*.fn.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00044000000000000007,
          0.00038500000000000003,
          0.00033
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "norm.weight"
        ],
        "label": "norm.weight"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Failedjointrecipe182namedLRsunchanged;9attentionNorm800to820;9FFNNorm840to820;NormWD0unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00022799999999999999,
          0.00019949999999999997,
          0.00017099999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "temperature_raw"
        ],
        "label": "temperature_raw"
      },
      {
        "lr": [
          0.000528,
          0.000462,
          0.000396
        ],
        "wd": 0.0,
        "names": [
          "candidate_variance_raw"
        ],
        "label": "candidate_variance_raw"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain"
        ],
        "label": "blocks.*.attention_log_gain"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight"
        ],
        "label": "blocks.*.an.weight, blocks.*.fn.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.2.qk.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.3.qk.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.4.qk.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.5.qk.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.6.qk.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.7.qk.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.8.qk.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.v.weight",
          "blocks.0.out.weight",
          "blocks.1.v.weight",
          "blocks.1.out.weight",
          "blocks.2.v.weight",
          "blocks.2.out.weight",
          "blocks.3.v.weight",
          "blocks.3.out.weight",
          "blocks.4.v.weight",
          "blocks.4.out.weight",
          "blocks.5.v.weight",
          "blocks.5.out.weight",
          "blocks.6.v.weight",
          "blocks.6.out.weight",
          "blocks.7.v.weight",
          "blocks.7.out.weight",
          "blocks.8.v.weight",
          "blocks.8.out.weight",
          "lm_head.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight"
        ],
        "label": "blocks.*.v.weight, blocks.*.out.weight, lm_head.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight"
      },
      {
        "lr": [
          0.0006084735202492212,
          0.0005324143302180685,
          0.0004563551401869158
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00078,
          0.0006825,
          0.0005849999999999999
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00044000000000000007,
          0.00038500000000000003,
          0.00033
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      },
      {
        "lr": [
          0.000552,
          0.000483,
          0.0004139999999999999
        ],
        "wd": 0.1,
        "names": [
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "norm.weight"
        ],
        "label": "norm.weight"
      },
      {
        "lr": [
          0.00045599999999999997,
          0.00039899999999999994,
          0.00034199999999999996
        ],
        "wd": 0.1,
        "names": [
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight"
        ],
        "label": "multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicitall200names:parent mos_branch_lr105 exceptlm_head .6to.63,mixing .6to.54"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00072,
          0.00063,
          0.00054
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00042,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0005764485981308411,
          0.000504392523364486,
          0.0004323364485981308
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          7.2e-05,
          6.3e-05,
          5.4000000000000005e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00043200000000000004,
          0.000378,
          0.000324
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00021600000000000002,
          0.000189,
          0.000162
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00072,
          0.00063,
          0.00054
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00028855813953488377,
          0.00025248837209302326,
          0.00021641860465116278
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00072,
          0.00063,
          0.00054
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00036,
          0.000315,
          0.00027
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.00072,
      0.00063,
      0.00054
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.00072,
        0.00063,
        0.00054
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Frozen parent384 LR role coefficients; only selected global/major/complement LR multiplied0.9; no extra width scaling"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Frozen parent384 LR role coefficients; only selected global/major/complement LR multiplied0.9; no extra width scaling"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00072,
          0.00063,
          0.00054
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005333333333333334,
          0.00046666666666666666,
          0.00039999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0005764485981308411,
          0.000504392523364486,
          0.00043233644859813077
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          7.200000000000002e-05,
          6.300000000000001e-05,
          5.4000000000000005e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00043200000000000004,
          0.000378,
          0.000324
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00021600000000000002,
          0.000189,
          0.000162
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00072,
          0.00063,
          0.00054
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00028855813953488377,
          0.00025248837209302326,
          0.00021641860465116278
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00072,
          0.00063,
          0.00054
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00036,
          0.000315,
          0.00027
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Frozen parent384 LR role coefficients; only selected global/major/complement LR multiplied0.9; no extra width scaling"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008800000000000001,
          0.0007700000000000001,
          0.00066
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005866666666666668,
          0.0005133333333333333,
          0.00043999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0007045482866043615,
          0.0006164797507788163,
          0.000528411214953271
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.800000000000002e-05,
          7.700000000000001e-05,
          6.6e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000528,
          0.000462,
          0.000396
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000264,
          0.000231,
          0.000198
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008800000000000001,
          0.0007700000000000001,
          0.00066
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00035268217054263573,
          0.00030859689922480623,
          0.00026451162790697673
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008800000000000001,
          0.0007700000000000001,
          0.00066
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00044000000000000007,
          0.00038500000000000003,
          0.00033
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008800000000000001,
      0.0007700000000000001,
      0.00066
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008800000000000001,
        0.0007700000000000001,
        0.00066
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Frozen parent384 LR role coefficients; only selected global/major/complement LR multiplied1.1; no extra width scaling"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005866666666666668,
          0.0005133333333333334,
          0.00044
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Frozen parent384 LR role coefficients; only selected global/major/complement LR multiplied1.1; no extra width scaling"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008800000000000001,
          0.0007700000000000001,
          0.00066
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005333333333333334,
          0.00046666666666666666,
          0.00039999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0007045482866043615,
          0.0006164797507788163,
          0.000528411214953271
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.800000000000001e-05,
          7.700000000000001e-05,
          6.6e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000528,
          0.000462,
          0.000396
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000264,
          0.000231,
          0.000198
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008800000000000001,
          0.0007700000000000001,
          0.00066
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00035268217054263573,
          0.00030859689922480623,
          0.00026451162790697673
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008800000000000001,
          0.0007700000000000001,
          0.00066
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00044000000000000007,
          0.00038500000000000003,
          0.00033
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Frozen parent384 LR role coefficients; only selected global/major/complement LR multiplied1.1; no extra width scaling"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005066666666666667,
          0.0004433333333333333,
          0.00037999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicit200namedLRs:unchanged384 major x0.95,other x1.0;numericWDsame;MoS includedmajor"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicit200namedLRs:unchanged384 major x0.9,other x1.05;numericWDsame;MoS includedmajor"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0004933333333333334,
          0.0004316666666666667,
          0.00037
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006565109034267912,
          0.0005744470404984422,
          0.0004923831775700934
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.2e-05,
          7.175e-05,
          6.149999999999999e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.0004919999999999999,
          0.0004304999999999999,
          0.0003689999999999999
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024599999999999996,
          0.00021524999999999995,
          0.00018449999999999996
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003286356589147287,
          0.00028755620155038755,
          0.0002464767441860465
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00041,
          0.00035874999999999997,
          0.00030749999999999994
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicit200namedLRs:unchanged384 major x0.925,other x1.025;numericWDsame;MoS includedmajor"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00045333333333333337,
          0.00039666666666666664,
          0.00033999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s082253a1be20c3",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Exact B200 finalLRs;103major x85/90;other97unchanged;relative to unchanged384 major0.85/other1.05;overrides applied once"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006084735202492212,
          0.0005324143302180685,
          0.0004563551401869158
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00022799999999999999,
          0.00019949999999999997,
          0.00017099999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "temperature_raw"
        ],
        "label": "temperature_raw"
      },
      {
        "lr": [
          0.000528,
          0.000462,
          0.000396
        ],
        "wd": 0.0,
        "names": [
          "candidate_variance_raw"
        ],
        "label": "candidate_variance_raw"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain"
        ],
        "label": "blocks.*.attention_log_gain"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.1.an.weight",
          "blocks.2.an.weight",
          "blocks.3.an.weight",
          "blocks.4.an.weight",
          "blocks.5.an.weight",
          "blocks.6.an.weight",
          "blocks.7.an.weight",
          "blocks.8.an.weight",
          "norm.weight"
        ],
        "label": "blocks.*.an.weight, norm.weight"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.fn.weight",
          "blocks.1.fn.weight",
          "blocks.2.fn.weight",
          "blocks.3.fn.weight",
          "blocks.4.fn.weight",
          "blocks.5.fn.weight",
          "blocks.6.fn.weight",
          "blocks.7.fn.weight",
          "blocks.8.fn.weight"
        ],
        "label": "blocks.*.fn.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00044000000000000007,
          0.00038500000000000003,
          0.00033
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "All103major sameB;other17roles fromapprovedproposal;67namedLR changes vsB;numericWD unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005066666666666667,
          0.0004433333333333333,
          0.00037999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Frozen384/1408 baseline actual role scales; major103 LR0.95, complement97 LR1.05; no extra width scaling"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005066666666666667,
          0.0004433333333333333,
          0.00037999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0005333333333333334,
          0.00046666666666666666,
          0.00039999999999999996
        ],
        "wd": 0.1,
        "names": [
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Preservejointmajor95/other105parent;restoreonly8MoSLRscales tooriginal384;allnumericWDunchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight",
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight",
          "blocks.9.attention_log_gain",
          "blocks.9.an.weight",
          "blocks.9.fn.weight",
          "blocks.9.candidate_reader.down.weight"
        ],
        "label": "embed.weight, blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight",
          "blocks.9.qk.weight",
          "blocks.9.v.weight",
          "blocks.9.attn_gate.proj.weight",
          "blocks.9.attn_gate.received.weight",
          "blocks.9.out.weight",
          "blocks.9.ffn.base.weight",
          "blocks.9.ffn.down.weight",
          "blocks.9.ffn.reader.aux1.weight",
          "blocks.9.ffn.reader.aux2.weight",
          "blocks.9.ffn.reader.router.query.weight",
          "blocks.9.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight",
          "blocks.9.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": null,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw",
          "blocks.9.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias",
          "blocks.9.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight",
          "blocks.9.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias",
          "blocks.9.candidate_reader.gate.weight",
          "blocks.9.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicit200namedLRs:unchanged384 major x0.9,other x1.05;numericWDsame;MoS includedmajor"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight",
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "embed.weight, blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": null,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicit200namedLRs:unchanged384 major x0.9,other x1.05;numericWDsame;MoS includedmajor"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight",
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "norm.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight"
        ],
        "label": "embed.weight, blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": null,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicit200namedLRs:unchanged384 major x0.9,other x1.05;numericWDsame;MoS includedmajor",
      "plasticity_parameter_lr_scale_overrides": {
        "heads.0.gate": 0.9,
        "heads.1.gate": 0.9,
        "heads.2.gate": 0.9,
        "heads.3.gate": 0.9,
        "heads.4.gate": 0.9,
        "heads.5.gate": 0.9,
        "heads.6.gate": 0.9,
        "heads.7.gate": 0.9
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight",
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "norm.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight"
        ],
        "label": "embed.weight, blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": null,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicit200namedLRs:unchanged384 major x0.9,other x1.05;numericWDsame;MoS includedmajor"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "base.embed.weight"
        ],
        "label": "base.embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "base.blocks.0.qk.weight",
          "base.blocks.0.v.weight",
          "base.blocks.0.attn_gate.proj.weight",
          "base.blocks.0.attn_gate.received.weight",
          "base.blocks.0.out.weight",
          "base.blocks.0.ffn.base.weight",
          "base.blocks.0.ffn.down.weight",
          "base.blocks.1.qk.weight",
          "base.blocks.1.v.weight",
          "base.blocks.1.attn_gate.proj.weight",
          "base.blocks.1.attn_gate.received.weight",
          "base.blocks.1.out.weight",
          "base.blocks.1.ffn.base.weight",
          "base.blocks.1.ffn.down.weight",
          "base.blocks.1.ffn.reader.aux1.weight",
          "base.blocks.1.ffn.reader.aux2.weight",
          "base.blocks.1.ffn.reader.router.query.weight",
          "base.blocks.1.ffn.reader.router.key.weight",
          "base.blocks.2.qk.weight",
          "base.blocks.2.v.weight",
          "base.blocks.2.attn_gate.proj.weight",
          "base.blocks.2.attn_gate.received.weight",
          "base.blocks.2.out.weight",
          "base.blocks.2.ffn.base.weight",
          "base.blocks.2.ffn.down.weight",
          "base.blocks.2.ffn.reader.aux1.weight",
          "base.blocks.2.ffn.reader.aux2.weight",
          "base.blocks.2.ffn.reader.router.query.weight",
          "base.blocks.2.ffn.reader.router.key.weight",
          "base.blocks.3.qk.weight",
          "base.blocks.3.v.weight",
          "base.blocks.3.attn_gate.proj.weight",
          "base.blocks.3.attn_gate.received.weight",
          "base.blocks.3.out.weight",
          "base.blocks.3.ffn.base.weight",
          "base.blocks.3.ffn.down.weight",
          "base.blocks.3.ffn.reader.aux1.weight",
          "base.blocks.3.ffn.reader.aux2.weight",
          "base.blocks.3.ffn.reader.router.query.weight",
          "base.blocks.3.ffn.reader.router.key.weight",
          "base.blocks.4.qk.weight",
          "base.blocks.4.v.weight",
          "base.blocks.4.attn_gate.proj.weight",
          "base.blocks.4.attn_gate.received.weight",
          "base.blocks.4.out.weight",
          "base.blocks.4.ffn.base.weight",
          "base.blocks.4.ffn.down.weight",
          "base.blocks.4.ffn.reader.aux1.weight",
          "base.blocks.4.ffn.reader.aux2.weight",
          "base.blocks.4.ffn.reader.router.query.weight",
          "base.blocks.4.ffn.reader.router.key.weight",
          "base.blocks.5.qk.weight",
          "base.blocks.5.v.weight",
          "base.blocks.5.attn_gate.proj.weight",
          "base.blocks.5.attn_gate.received.weight",
          "base.blocks.5.out.weight",
          "base.blocks.5.ffn.base.weight",
          "base.blocks.5.ffn.down.weight",
          "base.blocks.5.ffn.reader.aux1.weight",
          "base.blocks.5.ffn.reader.aux2.weight",
          "base.blocks.5.ffn.reader.router.query.weight",
          "base.blocks.5.ffn.reader.router.key.weight",
          "base.blocks.6.qk.weight",
          "base.blocks.6.v.weight",
          "base.blocks.6.attn_gate.proj.weight",
          "base.blocks.6.attn_gate.received.weight",
          "base.blocks.6.out.weight",
          "base.blocks.6.ffn.base.weight",
          "base.blocks.6.ffn.down.weight",
          "base.blocks.6.ffn.reader.aux1.weight",
          "base.blocks.6.ffn.reader.aux2.weight",
          "base.blocks.6.ffn.reader.router.query.weight",
          "base.blocks.6.ffn.reader.router.key.weight",
          "base.blocks.7.qk.weight",
          "base.blocks.7.v.weight",
          "base.blocks.7.attn_gate.proj.weight",
          "base.blocks.7.attn_gate.received.weight",
          "base.blocks.7.out.weight",
          "base.blocks.7.ffn.base.weight",
          "base.blocks.7.ffn.down.weight",
          "base.blocks.7.ffn.reader.aux1.weight",
          "base.blocks.7.ffn.reader.aux2.weight",
          "base.blocks.7.ffn.reader.router.query.weight",
          "base.blocks.7.ffn.reader.router.key.weight",
          "base.lm_head.weight",
          "base.multi_output.compress1.weight",
          "base.multi_output.compress2.weight",
          "base.multi_output.compress3.weight",
          "base.multi_output.facet1.weight",
          "base.multi_output.facet2.weight",
          "base.multi_output.facet3.weight",
          "base.multi_output.mixing.weight"
        ],
        "label": "base.blocks.*.qk.weight, base.blocks.*.v.weight, base.blocks.*.attn_gate.proj.weight, base.blocks.*.attn_gate.received.weight, base.blocks.*.out.weight, base.blocks.*.ffn.base.weight, base.blocks.*.ffn.down.weight, base.blocks.*.ffn.reader.aux1.weight, base.blocks.*.ffn.reader.aux2.weight, base.blocks.*.ffn.reader.router.query.weight, base.blocks.*.ffn.reader.router.key.weight, base.lm_head.weight, base.multi_output.compress1.weight, base.multi_output.compress2.weight, base.multi_output.compress3.weight, base.multi_output.facet1.weight, base.multi_output.facet2.weight, base.multi_output.facet3.weight, base.multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": 0.1,
        "names": [
          "base.blocks.0.ffn.spread_input.weight",
          "base.blocks.1.ffn.spread_input.weight",
          "base.blocks.2.ffn.spread_input.weight",
          "base.blocks.3.ffn.spread_input.weight",
          "base.blocks.4.ffn.spread_input.weight",
          "base.blocks.5.ffn.spread_input.weight",
          "base.blocks.6.ffn.spread_input.weight",
          "base.blocks.7.ffn.spread_input.weight"
        ],
        "label": "base.blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "base.rope_log_scale"
        ],
        "label": "base.rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "base.feedback_threshold_raw",
          "base.candidate_variance_raw",
          "base.blocks.0.ffn.variance_raw",
          "base.blocks.1.ffn.variance_raw",
          "base.blocks.2.ffn.variance_raw",
          "base.blocks.3.ffn.variance_raw",
          "base.blocks.4.ffn.variance_raw",
          "base.blocks.5.ffn.variance_raw",
          "base.blocks.6.ffn.variance_raw",
          "base.blocks.7.ffn.variance_raw"
        ],
        "label": "base.feedback_threshold_raw, base.candidate_variance_raw, base.blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": 0.0,
        "names": [
          "base.feedback_beta",
          "base.temperature_raw",
          "base.blocks.0.attn_gate.proj.bias",
          "base.blocks.1.attn_gate.proj.bias",
          "base.blocks.2.attn_gate.proj.bias",
          "base.blocks.3.attn_gate.proj.bias",
          "base.blocks.4.attn_gate.proj.bias",
          "base.blocks.5.attn_gate.proj.bias",
          "base.blocks.6.attn_gate.proj.bias",
          "base.blocks.7.attn_gate.proj.bias"
        ],
        "label": "base.feedback_beta, base.temperature_raw, base.blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.0,
        "names": [
          "base.blocks.0.attention_log_gain",
          "base.blocks.0.an.weight",
          "base.blocks.0.fn.weight",
          "base.blocks.1.attention_log_gain",
          "base.blocks.1.an.weight",
          "base.blocks.1.fn.weight",
          "base.blocks.2.attention_log_gain",
          "base.blocks.2.an.weight",
          "base.blocks.2.fn.weight",
          "base.blocks.3.attention_log_gain",
          "base.blocks.3.an.weight",
          "base.blocks.3.fn.weight",
          "base.blocks.4.attention_log_gain",
          "base.blocks.4.an.weight",
          "base.blocks.4.fn.weight",
          "base.blocks.5.attention_log_gain",
          "base.blocks.5.an.weight",
          "base.blocks.5.fn.weight",
          "base.blocks.6.attention_log_gain",
          "base.blocks.6.an.weight",
          "base.blocks.6.fn.weight",
          "base.blocks.7.attention_log_gain",
          "base.blocks.7.an.weight",
          "base.blocks.7.fn.weight",
          "base.norm.weight"
        ],
        "label": "base.blocks.*.attention_log_gain, base.blocks.*.an.weight, base.blocks.*.fn.weight, base.norm.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "base.blocks.0.candidate_reader.up.weight",
          "base.blocks.1.candidate_reader.up.weight",
          "base.blocks.2.candidate_reader.up.weight",
          "base.blocks.3.candidate_reader.up.weight",
          "base.blocks.4.candidate_reader.up.weight",
          "base.blocks.5.candidate_reader.up.weight",
          "base.blocks.6.candidate_reader.up.weight",
          "base.blocks.7.candidate_reader.up.weight"
        ],
        "label": "base.blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.05,
        "names": [
          "base.blocks.0.candidate_reader.down.weight",
          "base.blocks.1.candidate_reader.down.weight",
          "base.blocks.2.candidate_reader.down.weight",
          "base.blocks.3.candidate_reader.down.weight",
          "base.blocks.4.candidate_reader.down.weight",
          "base.blocks.5.candidate_reader.down.weight",
          "base.blocks.6.candidate_reader.down.weight",
          "base.blocks.7.candidate_reader.down.weight"
        ],
        "label": "base.blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": 0.0,
        "names": [
          "base.blocks.0.candidate_reader.gate.weight",
          "base.blocks.0.candidate_reader.gate.bias",
          "base.blocks.1.candidate_reader.gate.weight",
          "base.blocks.1.candidate_reader.gate.bias",
          "base.blocks.2.candidate_reader.gate.weight",
          "base.blocks.2.candidate_reader.gate.bias",
          "base.blocks.3.candidate_reader.gate.weight",
          "base.blocks.3.candidate_reader.gate.bias",
          "base.blocks.4.candidate_reader.gate.weight",
          "base.blocks.4.candidate_reader.gate.bias",
          "base.blocks.5.candidate_reader.gate.weight",
          "base.blocks.5.candidate_reader.gate.bias",
          "base.blocks.6.candidate_reader.gate.weight",
          "base.blocks.6.candidate_reader.gate.bias",
          "base.blocks.7.candidate_reader.gate.weight",
          "base.blocks.7.candidate_reader.gate.bias"
        ],
        "label": "base.blocks.*.candidate_reader.gate.weight, base.blocks.*.candidate_reader.gate.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "heads.0.key",
          "heads.0.readout",
          "heads.0.gate",
          "heads.0.value",
          "heads.0.rate_weight",
          "heads.1.key",
          "heads.1.readout",
          "heads.1.gate",
          "heads.1.value",
          "heads.1.rate_weight",
          "heads.2.key",
          "heads.2.readout",
          "heads.2.gate",
          "heads.2.value",
          "heads.2.rate_weight",
          "heads.3.key",
          "heads.3.readout",
          "heads.3.gate",
          "heads.3.value",
          "heads.3.rate_weight",
          "heads.4.key",
          "heads.4.readout",
          "heads.4.gate",
          "heads.4.value",
          "heads.4.rate_weight",
          "heads.5.key",
          "heads.5.readout",
          "heads.5.gate",
          "heads.5.value",
          "heads.5.rate_weight",
          "heads.6.key",
          "heads.6.readout",
          "heads.6.gate",
          "heads.6.value",
          "heads.6.rate_weight",
          "heads.7.key",
          "heads.7.readout",
          "heads.7.gate",
          "heads.7.value",
          "heads.7.rate_weight",
          "ffns.0.key",
          "ffns.0.readout",
          "ffns.0.value",
          "ffns.0.rate_weight",
          "ffns.1.key",
          "ffns.1.readout",
          "ffns.1.value",
          "ffns.1.rate_weight",
          "ffns.2.key",
          "ffns.2.readout",
          "ffns.2.value",
          "ffns.2.rate_weight",
          "ffns.3.key",
          "ffns.3.readout",
          "ffns.3.value",
          "ffns.3.rate_weight",
          "ffns.4.key",
          "ffns.4.readout",
          "ffns.4.value",
          "ffns.4.rate_weight",
          "ffns.5.key",
          "ffns.5.readout",
          "ffns.5.value",
          "ffns.5.rate_weight",
          "ffns.6.key",
          "ffns.6.readout",
          "ffns.6.value",
          "ffns.6.rate_weight",
          "ffns.7.key",
          "ffns.7.readout",
          "ffns.7.value",
          "ffns.7.rate_weight"
        ],
        "label": "heads.0.key, heads.0.readout, heads.0.gate, heads.0.value, heads.0.rate_weight, heads.1.key, heads.1.readout, heads.1.gate, heads.1.value, heads.1.rate_weight, heads.2.key, heads.2.readout, heads.2.gate, heads.2.value, heads.2.rate_weight, heads.3.key, heads.3.readout, heads.3.gate, heads.3.value, heads.3.rate_weight, heads.4.key, heads.4.readout, heads.4.gate, heads.4.value, heads.4.rate_weight, heads.5.key, heads.5.readout, heads.5.gate, heads.5.value, heads.5.rate_weight, heads.6.key, heads.6.readout, heads.6.gate, heads.6.value, heads.6.rate_weight, heads.7.key, heads.7.readout, heads.7.gate, heads.7.value, heads.7.rate_weight, ffns.0.key, ffns.0.readout, ffns.0.value, ffns.0.rate_weight, ffns.1.key, ffns.1.readout, ffns.1.value, ffns.1.rate_weight, ffns.2.key, ffns.2.readout, ffns.2.value, ffns.2.rate_weight, ffns.3.key, ffns.3.readout, ffns.3.value, ffns.3.rate_weight, ffns.4.key, ffns.4.readout, ffns.4.value, ffns.4.rate_weight, ffns.5.key, ffns.5.readout, ffns.5.value, ffns.5.rate_weight, ffns.6.key, ffns.6.readout, ffns.6.value, ffns.6.rate_weight, ffns.7.key, ffns.7.readout, ffns.7.value, ffns.7.rate_weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "heads.0.rate_bias",
          "heads.0.initial",
          "heads.1.rate_bias",
          "heads.1.initial",
          "heads.2.rate_bias",
          "heads.2.initial",
          "heads.3.rate_bias",
          "heads.3.initial",
          "heads.4.rate_bias",
          "heads.4.initial",
          "heads.5.rate_bias",
          "heads.5.initial",
          "heads.6.rate_bias",
          "heads.6.initial",
          "heads.7.rate_bias",
          "heads.7.initial",
          "ffns.0.rate_bias",
          "ffns.0.initial",
          "ffns.1.rate_bias",
          "ffns.1.initial",
          "ffns.2.rate_bias",
          "ffns.2.initial",
          "ffns.3.rate_bias",
          "ffns.3.initial",
          "ffns.4.rate_bias",
          "ffns.4.initial",
          "ffns.5.rate_bias",
          "ffns.5.initial",
          "ffns.6.rate_bias",
          "ffns.6.initial",
          "ffns.7.rate_bias",
          "ffns.7.initial"
        ],
        "label": "heads.0.rate_bias, heads.0.initial, heads.1.rate_bias, heads.1.initial, heads.2.rate_bias, heads.2.initial, heads.3.rate_bias, heads.3.initial, heads.4.rate_bias, heads.4.initial, heads.5.rate_bias, heads.5.initial, heads.6.rate_bias, heads.6.initial, heads.7.rate_bias, heads.7.initial, ffns.0.rate_bias, ffns.0.initial, ffns.1.rate_bias, ffns.1.initial, ffns.2.rate_bias, ffns.2.initial, ffns.3.rate_bias, ffns.3.initial, ffns.4.rate_bias, ffns.4.initial, ffns.5.rate_bias, ffns.5.initial, ffns.6.rate_bias, ffns.6.initial, ffns.7.rate_bias, ffns.7.initial"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicit200namedLRs:unchanged384 major x0.9,other x1.05;numericWDsame;MoS includedmajor"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight",
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight",
          "blocks.9.attention_log_gain",
          "blocks.9.an.weight",
          "blocks.9.fn.weight",
          "blocks.9.candidate_reader.down.weight"
        ],
        "label": "embed.weight, blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight",
          "blocks.9.qk.weight",
          "blocks.9.v.weight",
          "blocks.9.attn_gate.proj.weight",
          "blocks.9.attn_gate.received.weight",
          "blocks.9.out.weight",
          "blocks.9.ffn.base.weight",
          "blocks.9.ffn.down.weight",
          "blocks.9.ffn.reader.aux1.weight",
          "blocks.9.ffn.reader.aux2.weight",
          "blocks.9.ffn.reader.router.query.weight",
          "blocks.9.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight",
          "blocks.9.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": null,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw",
          "blocks.9.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias",
          "blocks.9.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight",
          "blocks.9.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias",
          "blocks.9.candidate_reader.gate.weight",
          "blocks.9.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "#3 exact named LR/WD retained for200 common parameters; newblock9 inherits block8 LR/WD; no depth LR scaling"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight",
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight",
          "blocks.9.attention_log_gain",
          "blocks.9.an.weight",
          "blocks.9.fn.weight",
          "blocks.9.candidate_reader.down.weight",
          "blocks.10.attention_log_gain",
          "blocks.10.an.weight",
          "blocks.10.fn.weight",
          "blocks.10.candidate_reader.down.weight"
        ],
        "label": "embed.weight, blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight",
          "blocks.9.qk.weight",
          "blocks.9.v.weight",
          "blocks.9.attn_gate.proj.weight",
          "blocks.9.attn_gate.received.weight",
          "blocks.9.out.weight",
          "blocks.9.ffn.base.weight",
          "blocks.9.ffn.down.weight",
          "blocks.9.ffn.reader.aux1.weight",
          "blocks.9.ffn.reader.aux2.weight",
          "blocks.9.ffn.reader.router.query.weight",
          "blocks.9.ffn.reader.router.key.weight",
          "blocks.10.qk.weight",
          "blocks.10.v.weight",
          "blocks.10.attn_gate.proj.weight",
          "blocks.10.attn_gate.received.weight",
          "blocks.10.out.weight",
          "blocks.10.ffn.base.weight",
          "blocks.10.ffn.down.weight",
          "blocks.10.ffn.reader.aux1.weight",
          "blocks.10.ffn.reader.aux2.weight",
          "blocks.10.ffn.reader.router.query.weight",
          "blocks.10.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight",
          "blocks.9.ffn.spread_input.weight",
          "blocks.10.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": null,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw",
          "blocks.9.ffn.variance_raw",
          "blocks.10.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias",
          "blocks.9.attn_gate.proj.bias",
          "blocks.10.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight",
          "blocks.9.candidate_reader.up.weight",
          "blocks.10.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias",
          "blocks.9.candidate_reader.gate.weight",
          "blocks.9.candidate_reader.gate.bias",
          "blocks.10.candidate_reader.gate.weight",
          "blocks.10.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "#3 all200 named LR/WD retained; addedblocks9/10 inherit block8 exact role LR/WD"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight",
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight",
          "blocks.9.attention_log_gain",
          "blocks.9.an.weight",
          "blocks.9.fn.weight",
          "blocks.9.candidate_reader.down.weight",
          "blocks.10.attention_log_gain",
          "blocks.10.an.weight",
          "blocks.10.fn.weight",
          "blocks.10.candidate_reader.down.weight",
          "blocks.11.attention_log_gain",
          "blocks.11.an.weight",
          "blocks.11.fn.weight",
          "blocks.11.candidate_reader.down.weight"
        ],
        "label": "embed.weight, blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight",
          "blocks.9.qk.weight",
          "blocks.9.v.weight",
          "blocks.9.attn_gate.proj.weight",
          "blocks.9.attn_gate.received.weight",
          "blocks.9.out.weight",
          "blocks.9.ffn.base.weight",
          "blocks.9.ffn.down.weight",
          "blocks.9.ffn.reader.aux1.weight",
          "blocks.9.ffn.reader.aux2.weight",
          "blocks.9.ffn.reader.router.query.weight",
          "blocks.9.ffn.reader.router.key.weight",
          "blocks.10.qk.weight",
          "blocks.10.v.weight",
          "blocks.10.attn_gate.proj.weight",
          "blocks.10.attn_gate.received.weight",
          "blocks.10.out.weight",
          "blocks.10.ffn.base.weight",
          "blocks.10.ffn.down.weight",
          "blocks.10.ffn.reader.aux1.weight",
          "blocks.10.ffn.reader.aux2.weight",
          "blocks.10.ffn.reader.router.query.weight",
          "blocks.10.ffn.reader.router.key.weight",
          "blocks.11.qk.weight",
          "blocks.11.v.weight",
          "blocks.11.attn_gate.proj.weight",
          "blocks.11.attn_gate.received.weight",
          "blocks.11.out.weight",
          "blocks.11.ffn.base.weight",
          "blocks.11.ffn.down.weight",
          "blocks.11.ffn.reader.aux1.weight",
          "blocks.11.ffn.reader.aux2.weight",
          "blocks.11.ffn.reader.router.query.weight",
          "blocks.11.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight",
          "blocks.9.ffn.spread_input.weight",
          "blocks.10.ffn.spread_input.weight",
          "blocks.11.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": null,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw",
          "blocks.9.ffn.variance_raw",
          "blocks.10.ffn.variance_raw",
          "blocks.11.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias",
          "blocks.9.attn_gate.proj.bias",
          "blocks.10.attn_gate.proj.bias",
          "blocks.11.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight",
          "blocks.9.candidate_reader.up.weight",
          "blocks.10.candidate_reader.up.weight",
          "blocks.11.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias",
          "blocks.9.candidate_reader.gate.weight",
          "blocks.9.candidate_reader.gate.bias",
          "blocks.10.candidate_reader.gate.weight",
          "blocks.10.candidate_reader.gate.bias",
          "blocks.11.candidate_reader.gate.weight",
          "blocks.11.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "#3 all200 named LR/WD retained; addedblocks9/10 inherit block8 exact role LR/WD"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight",
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "norm.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight"
        ],
        "label": "embed.weight, blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": null,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "#3 all200 named LR/WD retained; addedblocks9/10 inherit block8 exact role LR/WD"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.2.qk.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.3.qk.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.4.qk.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.5.qk.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.6.qk.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.7.qk.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.8.qk.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "lm_head.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, lm_head.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.v.weight",
          "blocks.0.out.weight",
          "blocks.1.v.weight",
          "blocks.1.out.weight",
          "blocks.2.v.weight",
          "blocks.2.out.weight",
          "blocks.3.v.weight",
          "blocks.3.out.weight",
          "blocks.4.v.weight",
          "blocks.4.out.weight",
          "blocks.5.v.weight",
          "blocks.5.out.weight",
          "blocks.6.v.weight",
          "blocks.6.out.weight",
          "blocks.7.v.weight",
          "blocks.7.out.weight",
          "blocks.8.v.weight",
          "blocks.8.out.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight"
        ],
        "label": "blocks.*.v.weight, blocks.*.out.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight"
      },
      {
        "lr": [
          0.0006084735202492212,
          0.0005324143302180685,
          0.0004563551401869158
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.000552,
          0.000483,
          0.0004139999999999999
        ],
        "wd": 0.1,
        "names": [
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00022799999999999999,
          0.00019949999999999997,
          0.00017099999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "temperature_raw"
        ],
        "label": "temperature_raw"
      },
      {
        "lr": [
          0.000528,
          0.000462,
          0.000396
        ],
        "wd": 0.0,
        "names": [
          "candidate_variance_raw"
        ],
        "label": "candidate_variance_raw"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain"
        ],
        "label": "blocks.*.attention_log_gain"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight"
        ],
        "label": "blocks.*.an.weight, blocks.*.fn.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00078,
          0.0006825,
          0.0005849999999999999
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00044000000000000007,
          0.00038500000000000003,
          0.00033
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "norm.weight"
        ],
        "label": "norm.weight"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicit all200 named multipliers inherited from jointAttentionReader parent except six MoS compress/facet .6 to .63"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0005333333333333334,
          0.00046666666666666666,
          0.00039999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight"
      },
      {
        "lr": [
          0.0005066666666666667,
          0.0004433333333333333,
          0.00037999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight"
        ],
        "label": "blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight"
      },
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Preserveallparentinput/outputmatchedLRs;onlyMoS8matrixLRs x0.95;allnumericWDfixed"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006084735202492212,
          0.0005324143302180685,
          0.0004563551401869158
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00022799999999999999,
          0.00019949999999999997,
          0.00017099999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "temperature_raw"
        ],
        "label": "temperature_raw"
      },
      {
        "lr": [
          0.000528,
          0.000462,
          0.000396
        ],
        "wd": 0.0,
        "names": [
          "candidate_variance_raw"
        ],
        "label": "candidate_variance_raw"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain"
        ],
        "label": "blocks.*.attention_log_gain"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight"
        ],
        "label": "blocks.*.an.weight, blocks.*.fn.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "norm.weight"
        ],
        "label": "norm.weight"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "OriginalNorm820173namedLRsunchanged;9Readerdown760to800and18Readergate440to420micro;allnumericWDunchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676012,
          0.000560436137071651,
          0.0004803738317757008
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00022799999999999999,
          0.00019949999999999997,
          0.00017099999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "temperature_raw"
        ],
        "label": "temperature_raw"
      },
      {
        "lr": [
          0.000528,
          0.000462,
          0.000396
        ],
        "wd": 0.0,
        "names": [
          "candidate_variance_raw"
        ],
        "label": "candidate_variance_raw"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain"
        ],
        "label": "blocks.*.attention_log_gain"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight"
        ],
        "label": "blocks.*.an.weight, blocks.*.fn.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00044000000000000007,
          0.00038500000000000003,
          0.00033
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "norm.weight"
        ],
        "label": "norm.weight"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Norm820parent191namedLRsunchanged;9spread_inputmatrixLRs+5.2631579%;allnumericWDunchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0005780498442367601,
          0.0005057936137071651,
          0.00043353738317757003
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00022799999999999999,
          0.00019949999999999997,
          0.00017099999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "temperature_raw"
        ],
        "label": "temperature_raw"
      },
      {
        "lr": [
          0.000528,
          0.000462,
          0.000396
        ],
        "wd": 0.0,
        "names": [
          "candidate_variance_raw"
        ],
        "label": "candidate_variance_raw"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain"
        ],
        "label": "blocks.*.attention_log_gain"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight"
        ],
        "label": "blocks.*.an.weight, blocks.*.fn.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00044000000000000007,
          0.00038500000000000003,
          0.00033
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "norm.weight"
        ],
        "label": "norm.weight"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "OriginalNorm820parent191LRsunchanged;9spread_inputmatrixLRs*0.95;allnumericWDunchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.2.qk.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.3.qk.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.4.qk.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.5.qk.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.6.qk.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.7.qk.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.8.qk.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.v.weight",
          "blocks.0.out.weight",
          "blocks.1.v.weight",
          "blocks.1.out.weight",
          "blocks.2.v.weight",
          "blocks.2.out.weight",
          "blocks.3.v.weight",
          "blocks.3.out.weight",
          "blocks.4.v.weight",
          "blocks.4.out.weight",
          "blocks.5.v.weight",
          "blocks.5.out.weight",
          "blocks.6.v.weight",
          "blocks.6.out.weight",
          "blocks.7.v.weight",
          "blocks.7.out.weight",
          "blocks.8.v.weight",
          "blocks.8.out.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight"
        ],
        "label": "blocks.*.v.weight, blocks.*.out.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight"
      },
      {
        "lr": [
          0.0006084735202492212,
          0.0005324143302180685,
          0.0004563551401869158
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.000552,
          0.000483,
          0.0004139999999999999
        ],
        "wd": 0.1,
        "names": [
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight"
      },
      {
        "lr": [
          0.00043200000000000004,
          0.000378,
          0.000324
        ],
        "wd": 0.1,
        "names": [
          "multi_output.mixing.weight"
        ],
        "label": "multi_output.mixing.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00022799999999999999,
          0.00019949999999999997,
          0.00017099999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "temperature_raw"
        ],
        "label": "temperature_raw"
      },
      {
        "lr": [
          0.000528,
          0.000462,
          0.000396
        ],
        "wd": 0.0,
        "names": [
          "candidate_variance_raw"
        ],
        "label": "candidate_variance_raw"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain"
        ],
        "label": "blocks.*.attention_log_gain"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight"
        ],
        "label": "blocks.*.an.weight, blocks.*.fn.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00078,
          0.0006825,
          0.0005849999999999999
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00044000000000000007,
          0.00038500000000000003,
          0.00033
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "norm.weight"
        ],
        "label": "norm.weight"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicitall200names:parent mos_branch_lr105 exceptlm_head .6to.63,mixing .6to.54"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006084735202492212,
          0.0005324143302180685,
          0.0004563551401869158
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.000552,
          0.000483,
          0.0004139999999999999
        ],
        "wd": 0.1,
        "names": [
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00022799999999999999,
          0.00019949999999999997,
          0.00017099999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta"
        ],
        "label": "feedback_beta"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "temperature_raw"
        ],
        "label": "temperature_raw"
      },
      {
        "lr": [
          0.000528,
          0.000462,
          0.000396
        ],
        "wd": 0.0,
        "names": [
          "candidate_variance_raw"
        ],
        "label": "candidate_variance_raw"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain"
        ],
        "label": "blocks.*.attention_log_gain"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight"
        ],
        "label": "blocks.*.an.weight, blocks.*.fn.weight"
      },
      {
        "lr": [
          0.0002394,
          0.00020947499999999998,
          0.00017954999999999997
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00078,
          0.0006825,
          0.0005849999999999999
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00044000000000000007,
          0.00038500000000000003,
          0.00033
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "norm.weight"
        ],
        "label": "norm.weight"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicit200namedLRs;36QK/received-gate parameters x1.05 vs jointAttentionReader parent;164LR and all numericWD unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006084735202492212,
          0.0005324143302180685,
          0.0004563551401869158
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0006000000000000001,
          0.000525,
          0.00045
        ],
        "wd": 0.1,
        "names": [
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00022799999999999999,
          0.00019949999999999997,
          0.00017099999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "temperature_raw"
        ],
        "label": "temperature_raw"
      },
      {
        "lr": [
          0.000528,
          0.000462,
          0.000396
        ],
        "wd": 0.0,
        "names": [
          "candidate_variance_raw"
        ],
        "label": "candidate_variance_raw"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain"
        ],
        "label": "blocks.*.attention_log_gain"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight"
        ],
        "label": "blocks.*.an.weight, blocks.*.fn.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "norm.weight"
        ],
        "label": "norm.weight"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "16DeltaRouterquery/key600to750baseLRratio;parentReader800gate420other184namedLRsandallWDunchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.2.qk.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.3.qk.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.4.qk.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.5.qk.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.6.qk.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.7.qk.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.8.qk.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.v.weight",
          "blocks.0.out.weight",
          "blocks.1.v.weight",
          "blocks.1.out.weight",
          "blocks.2.v.weight",
          "blocks.2.out.weight",
          "blocks.3.v.weight",
          "blocks.3.out.weight",
          "blocks.4.v.weight",
          "blocks.4.out.weight",
          "blocks.5.v.weight",
          "blocks.5.out.weight",
          "blocks.6.v.weight",
          "blocks.6.out.weight",
          "blocks.7.v.weight",
          "blocks.7.out.weight",
          "blocks.8.v.weight",
          "blocks.8.out.weight",
          "lm_head.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight"
        ],
        "label": "blocks.*.v.weight, blocks.*.out.weight, lm_head.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight"
      },
      {
        "lr": [
          0.0006084735202492212,
          0.0005324143302180685,
          0.0004563551401869158
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.000552,
          0.000483,
          0.0004139999999999999
        ],
        "wd": 0.1,
        "names": [
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight"
        ],
        "label": "blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight"
      },
      {
        "lr": [
          0.00045599999999999997,
          0.00039899999999999994,
          0.00034199999999999996
        ],
        "wd": 0.1,
        "names": [
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight"
        ],
        "label": "multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00022799999999999999,
          0.00019949999999999997,
          0.00017099999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "temperature_raw"
        ],
        "label": "temperature_raw"
      },
      {
        "lr": [
          0.000528,
          0.000462,
          0.000396
        ],
        "wd": 0.0,
        "names": [
          "candidate_variance_raw"
        ],
        "label": "candidate_variance_raw"
      },
      {
        "lr": [
          0.00076,
          0.000665,
          0.00057
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.1.attention_log_gain",
          "blocks.2.attention_log_gain",
          "blocks.3.attention_log_gain",
          "blocks.4.attention_log_gain",
          "blocks.5.attention_log_gain",
          "blocks.6.attention_log_gain",
          "blocks.7.attention_log_gain",
          "blocks.8.attention_log_gain"
        ],
        "label": "blocks.*.attention_log_gain"
      },
      {
        "lr": [
          0.00082,
          0.0007174999999999999,
          0.0006149999999999999
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.an.weight",
          "blocks.8.fn.weight"
        ],
        "label": "blocks.*.an.weight, blocks.*.fn.weight"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00078,
          0.0006825,
          0.0005849999999999999
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00044000000000000007,
          0.00038500000000000003,
          0.00033
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "norm.weight"
        ],
        "label": "norm.weight"
      }
    ],
    "basis": "source:sb18f63949d7bec",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Explicit all200: parent except compress1/2/3 .63to.57; mixing .54to.60"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight",
          "multi_output.prediction_gate.up.bias",
          "multi_output.prediction_gate.down.bias"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, multi_output.prediction_gate.up.bias, multi_output.prediction_gate.down.bias"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight",
          "multi_output.prediction_gate.up.weight",
          "multi_output.prediction_gate.down.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight, multi_output.prediction_gate.up.weight, multi_output.prediction_gate.down.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sa10fdfd5c25eff",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Parent320 final named LR/WD unchanged; new gate matrices0.8x WD0.1, biases1x WD0; native gate fan-in initialization separately managed",
      "effective_parameter_policy": {
        "mode": "frozen_named_parent",
        "profile": "references/extended320_OPTIMIZER_PARAMETER_RULES.json",
        "sha256": "61ed496a87bfbea9b48d8de7caec57ef3b69af7911538c54ff0ae8e645edee91",
        "parent_contract": "references/parent320_CONTRACT.json",
        "parent_contract_sha256": "865064dc227a1e089d5ccf170b3b94256a6c42b0bb57525c367a38811d582b46",
        "target_geometry": {
          "width": 320,
          "hidden": 1120,
          "layers": 9,
          "q_heads": 10,
          "head_dim": 32,
          "reader_width": 48,
          "mixture_gate_hidden": 320
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "All200 shared parameters exactly initialized as parent320; new gate up native fan-in normal, all biases zero, down matrix zero. Gate allocation uses isolated RNG. No pretrained checkpoint.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw",
          "blocks.9.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias",
          "blocks.9.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "blocks.9.attention_log_gain",
          "blocks.9.an.weight",
          "blocks.9.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "blocks.9.qk.weight",
          "blocks.9.v.weight",
          "blocks.9.attn_gate.proj.weight",
          "blocks.9.attn_gate.received.weight",
          "blocks.9.out.weight",
          "blocks.9.ffn.base.weight",
          "blocks.9.ffn.down.weight",
          "blocks.9.ffn.reader.aux1.weight",
          "blocks.9.ffn.reader.aux2.weight",
          "blocks.9.ffn.reader.router.query.weight",
          "blocks.9.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight",
          "blocks.9.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight",
          "blocks.9.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight",
          "blocks.9.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias",
          "blocks.9.candidate_reader.gate.weight",
          "blocks.9.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s41fd6c43ee10ff",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Frozen parent320 named final LR/WD; added block9 inherits corresponding block8 coefficient. No depth or width recomputation.",
      "effective_parameter_policy": {
        "mode": "frozen_named_parent",
        "profile": "references/extended320_OPTIMIZER_PARAMETER_RULES.json",
        "sha256": "ef37d4d2e6286059783127a27993d3acc3b8ebc1dc58488be1a28767ddc2a580",
        "parent_contract": "references/parent320_CONTRACT.json",
        "parent_contract_sha256": "865064dc227a1e089d5ccf170b3b94256a6c42b0bb57525c367a38811d582b46",
        "target_geometry": {
          "layers": 10,
          "width": 320,
          "hidden": 1120,
          "q_heads": 10,
          "head_dim": 32,
          "initialization_layers": 9
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh0M; parent320 common parameter initialization exactly preserved; new tenth block has independent name-derived seed and same initialization coefficients; depth scaling reference fixed9.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; actual10-layer speed and capacity measured before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0005333333333333334,
          0.00046666666666666666,
          0.00039999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.down.weight",
          "blocks.1.ffn.down.weight",
          "blocks.2.ffn.down.weight",
          "blocks.3.ffn.down.weight",
          "blocks.4.ffn.down.weight",
          "blocks.5.ffn.down.weight",
          "blocks.6.ffn.down.weight",
          "blocks.7.ffn.down.weight",
          "blocks.8.ffn.down.weight"
        ],
        "label": "blocks.*.ffn.down.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s837baf1d187ad9",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent fresh named-parameter RNG initialization; hidden1120→1344; down std ratio sqrt(1120/1344),down LR ratio1120/1344; unchanged-shape parameters identical to parent fresh320; no trained weights or initial checkpoint files.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s731bc35428cd9e",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Copy final lr_scale and weight_decay per parameter from frozen best320 profile. Never multiply role or target width ratios again.",
      "effective_parameter_policy": {
        "mode": "frozen_named_parent",
        "profile": "references/parent320_OPTIMIZER_PARAMETER_RULES.json",
        "sha256": "adfb05fb57ebf24b316d477c5200d817b4c61937dafe9c263f42fff9899bd43e",
        "parent_contract": "references/parent320_CONTRACT.json",
        "parent_contract_sha256": "865064dc227a1e089d5ccf170b3b94256a6c42b0bb57525c367a38811d582b46",
        "target_geometry": {
          "width": 384,
          "hidden": 1344,
          "layers": 9,
          "q_heads": 12,
          "kv_heads": 12,
          "head_dim": 32,
          "reader_width": 48,
          "output_components": 3,
          "output_state_layers": 3
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh384 with fan-in initialization ONLY, reference width256/FFN896. Native branch compression; facet h9 identity retained, context columns fan-in scaled. Existing untrained256 fixture for unchanged shapes and RNG. LR/WD copied independently from saved320 optimizer table.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight",
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "embed.weight, blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": null,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Copy#3 final200named actual coefficients directly; no second width/fan-in multiplier. All other original#1/#2 learning semantics retained."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh320 native fan-in branch compression; C3 copies C2 into h8/h9 columns and zero h7 columns, then all columns learn. Preserve facet h9 identity; width-scale context half. Existing untrained256 zero fixture only for unchanged-shape parameters/RNG in RAM. Fresh Adam; no new initial/recovery weight files.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00084,
          0.000735,
          0.0006299999999999999
        ],
        "wd": null,
        "names": [
          "embed.weight",
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight",
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "embed.weight, blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight, blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": null,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight",
          "multi_output.compress.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight, multi_output.compress.weight"
      },
      {
        "lr": [
          0.0006725233644859813,
          0.0005884579439252336,
          0.000504392523364486
        ],
        "wd": null,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          8.400000000000001e-05,
          7.350000000000001e-05,
          6.3e-05
        ],
        "wd": null,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.000504,
          0.000441,
          0.00037799999999999997
        ],
        "wd": null,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.000252,
          0.0002205,
          0.00018899999999999999
        ],
        "wd": null,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00033665116279069774,
          0.0002945697674418605,
          0.00025248837209302326
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.00042,
          0.0003675,
          0.00031499999999999996
        ],
        "wd": null,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_lr_scale_overrides（WD未復元）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Copy#3 final coefficients for197same-named tensors; shared MoS compression receives same0.6coefficient as all3branch compression matrices in#3. No extra width multiplier."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "New0M native per-name seed, explicit width initialization, same-shape parent zero parameters, parent initial training RNG; fresh Adam",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qkv.weight",
          "blocks.0.attn_out.weight",
          "blocks.0.gate_up.weight",
          "blocks.0.down.weight",
          "blocks.1.d_write.context_key.weight",
          "blocks.1.d_write.delta_key.0.weight",
          "blocks.1.d_write.delta_key.1.weight",
          "blocks.1.d_write.value.0.weight",
          "blocks.1.d_write.value.1.weight",
          "blocks.1.d_read.query.weight",
          "blocks.1.d_read.out.weight",
          "blocks.1.qkv.weight",
          "blocks.1.attn_out.weight",
          "blocks.1.gate_up.weight",
          "blocks.1.down.weight",
          "blocks.2.d_write.context_key.weight",
          "blocks.2.d_write.delta_key.0.weight",
          "blocks.2.d_write.delta_key.1.weight",
          "blocks.2.d_write.value.0.weight",
          "blocks.2.d_write.value.1.weight",
          "blocks.2.d_read.query.weight",
          "blocks.2.d_read.out.weight",
          "blocks.2.qkv.weight",
          "blocks.2.attn_out.weight",
          "blocks.2.gate_up.weight",
          "blocks.2.down.weight",
          "blocks.3.d_write.context_key.weight",
          "blocks.3.d_write.delta_key.0.weight",
          "blocks.3.d_write.delta_key.1.weight",
          "blocks.3.d_write.value.0.weight",
          "blocks.3.d_write.value.1.weight",
          "blocks.3.d_read.query.weight",
          "blocks.3.d_read.out.weight",
          "blocks.3.qkv.weight",
          "blocks.3.attn_out.weight",
          "blocks.3.gate_up.weight",
          "blocks.3.down.weight",
          "blocks.4.d_write.context_key.weight",
          "blocks.4.d_write.delta_key.0.weight",
          "blocks.4.d_write.delta_key.1.weight",
          "blocks.4.d_write.value.0.weight",
          "blocks.4.d_write.value.1.weight",
          "blocks.4.d_read.query.weight",
          "blocks.4.d_read.out.weight",
          "blocks.4.qkv.weight",
          "blocks.4.attn_out.weight",
          "blocks.4.gate_up.weight",
          "blocks.4.down.weight",
          "blocks.5.d_write.context_key.weight",
          "blocks.5.d_write.delta_key.0.weight",
          "blocks.5.d_write.delta_key.1.weight",
          "blocks.5.d_write.value.0.weight",
          "blocks.5.d_write.value.1.weight",
          "blocks.5.d_read.query.weight",
          "blocks.5.d_read.out.weight",
          "blocks.5.qkv.weight",
          "blocks.5.attn_out.weight",
          "blocks.5.gate_up.weight",
          "blocks.5.down.weight",
          "blocks.6.d_write.context_key.weight",
          "blocks.6.d_write.delta_key.0.weight",
          "blocks.6.d_write.delta_key.1.weight",
          "blocks.6.d_write.value.0.weight",
          "blocks.6.d_write.value.1.weight",
          "blocks.6.d_read.query.weight",
          "blocks.6.d_read.out.weight",
          "blocks.6.qkv.weight",
          "blocks.6.attn_out.weight",
          "blocks.6.gate_up.weight",
          "blocks.6.down.weight",
          "blocks.7.d_write.context_key.weight",
          "blocks.7.d_write.delta_key.0.weight",
          "blocks.7.d_write.delta_key.1.weight",
          "blocks.7.d_write.value.0.weight",
          "blocks.7.d_write.value.1.weight",
          "blocks.7.d_read.query.weight",
          "blocks.7.d_read.out.weight",
          "blocks.7.qkv.weight",
          "blocks.7.attn_out.weight",
          "blocks.7.gate_up.weight",
          "blocks.7.down.weight",
          "embedding.weight",
          "head.weight"
        ],
        "label": "blocks.*.qkv.weight, blocks.*.attn_out.weight, blocks.*.gate_up.weight, blocks.*.down.weight, blocks.*.d_write.context_key.weight, blocks.*.d_write.delta_key.0.weight, blocks.*.d_write.delta_key.1.weight, blocks.*.d_write.value.0.weight, blocks.*.d_write.value.1.weight, blocks.*.d_read.query.weight, blocks.*.d_read.out.weight, embedding.weight, head.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.raw_gamma",
          "blocks.0.a_res.raw_eta",
          "blocks.0.f_res.raw_eta",
          "blocks.1.raw_gamma",
          "blocks.1.d_write.retention.proj.weight",
          "blocks.1.d_write.retention.proj.bias",
          "blocks.1.d_write.write.proj.weight",
          "blocks.1.d_write.write.proj.bias",
          "blocks.1.d_write.allocation.weight",
          "blocks.1.d_write.allocation.bias",
          "blocks.1.d_write.dual_selector.weight",
          "blocks.1.d_write.dual_selector.bias",
          "blocks.1.d_write.dual_strength.weight",
          "blocks.1.d_write.dual_strength.bias",
          "blocks.1.d_read.residual.raw_eta",
          "blocks.1.a_res.raw_eta",
          "blocks.1.f_res.raw_eta",
          "blocks.2.raw_gamma",
          "blocks.2.d_write.retention.proj.weight",
          "blocks.2.d_write.retention.proj.bias",
          "blocks.2.d_write.write.proj.weight",
          "blocks.2.d_write.write.proj.bias",
          "blocks.2.d_write.allocation.weight",
          "blocks.2.d_write.allocation.bias",
          "blocks.2.d_write.dual_selector.weight",
          "blocks.2.d_write.dual_selector.bias",
          "blocks.2.d_write.dual_strength.weight",
          "blocks.2.d_write.dual_strength.bias",
          "blocks.2.d_read.residual.raw_eta",
          "blocks.2.a_res.raw_eta",
          "blocks.2.f_res.raw_eta",
          "blocks.3.raw_gamma",
          "blocks.3.d_write.retention.proj.weight",
          "blocks.3.d_write.retention.proj.bias",
          "blocks.3.d_write.write.proj.weight",
          "blocks.3.d_write.write.proj.bias",
          "blocks.3.d_write.allocation.weight",
          "blocks.3.d_write.allocation.bias",
          "blocks.3.d_write.dual_selector.weight",
          "blocks.3.d_write.dual_selector.bias",
          "blocks.3.d_write.dual_strength.weight",
          "blocks.3.d_write.dual_strength.bias",
          "blocks.3.d_read.residual.raw_eta",
          "blocks.3.a_res.raw_eta",
          "blocks.3.f_res.raw_eta",
          "blocks.4.raw_gamma",
          "blocks.4.d_write.retention.proj.weight",
          "blocks.4.d_write.retention.proj.bias",
          "blocks.4.d_write.write.proj.weight",
          "blocks.4.d_write.write.proj.bias",
          "blocks.4.d_write.allocation.weight",
          "blocks.4.d_write.allocation.bias",
          "blocks.4.d_write.dual_selector.weight",
          "blocks.4.d_write.dual_selector.bias",
          "blocks.4.d_write.dual_strength.weight",
          "blocks.4.d_write.dual_strength.bias",
          "blocks.4.d_read.residual.raw_eta",
          "blocks.4.a_res.raw_eta",
          "blocks.4.f_res.raw_eta",
          "blocks.5.raw_gamma",
          "blocks.5.d_write.retention.proj.weight",
          "blocks.5.d_write.retention.proj.bias",
          "blocks.5.d_write.write.proj.weight",
          "blocks.5.d_write.write.proj.bias",
          "blocks.5.d_write.allocation.weight",
          "blocks.5.d_write.allocation.bias",
          "blocks.5.d_write.dual_selector.weight",
          "blocks.5.d_write.dual_selector.bias",
          "blocks.5.d_write.dual_strength.weight",
          "blocks.5.d_write.dual_strength.bias",
          "blocks.5.d_read.residual.raw_eta",
          "blocks.5.a_res.raw_eta",
          "blocks.5.f_res.raw_eta",
          "blocks.6.raw_gamma",
          "blocks.6.d_write.retention.proj.weight",
          "blocks.6.d_write.retention.proj.bias",
          "blocks.6.d_write.write.proj.weight",
          "blocks.6.d_write.write.proj.bias",
          "blocks.6.d_write.allocation.weight",
          "blocks.6.d_write.allocation.bias",
          "blocks.6.d_write.dual_selector.weight",
          "blocks.6.d_write.dual_selector.bias",
          "blocks.6.d_write.dual_strength.weight",
          "blocks.6.d_write.dual_strength.bias",
          "blocks.6.d_read.residual.raw_eta",
          "blocks.6.a_res.raw_eta",
          "blocks.6.f_res.raw_eta",
          "blocks.7.raw_gamma",
          "blocks.7.d_write.retention.proj.weight",
          "blocks.7.d_write.retention.proj.bias",
          "blocks.7.d_write.write.proj.weight",
          "blocks.7.d_write.write.proj.bias",
          "blocks.7.d_write.allocation.weight",
          "blocks.7.d_write.allocation.bias",
          "blocks.7.d_write.dual_selector.weight",
          "blocks.7.d_write.dual_selector.bias",
          "blocks.7.d_write.dual_strength.weight",
          "blocks.7.d_write.dual_strength.bias",
          "blocks.7.d_read.residual.raw_eta",
          "blocks.7.a_res.raw_eta",
          "blocks.7.f_res.raw_eta"
        ],
        "label": "blocks.*.raw_gamma, blocks.*.a_res.raw_eta, blocks.*.f_res.raw_eta, blocks.*.d_write.retention.proj.weight, blocks.*.d_write.retention.proj.bias, blocks.*.d_write.write.proj.weight, blocks.*.d_write.write.proj.bias, blocks.*.d_write.allocation.weight, blocks.*.d_write.allocation.bias, blocks.*.d_write.dual_selector.weight, blocks.*.d_write.dual_selector.bias, blocks.*.d_write.dual_strength.weight, blocks.*.d_write.dual_strength.bias, blocks.*.d_read.residual.raw_eta"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0004,
      0.00035,
      0.0003
    ],
    "schedule": {
      "lr": [
        0.0004,
        0.00035,
        0.0003
      ],
      "start_words": [
        0,
        30000000,
        60000000
      ],
      "warmup_words": 1600000
    },
    "optimizer": {
      "type": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "mean supervised-token causal NLL; matched logical-window and group masks",
    "initialization": "all156 tensors exact #298",
    "runtime_optimizations": [
      "FP32 recompute D",
      "fused compiled body",
      "padded microbatch budget8192",
      "compile cache64"
    ]
  },
  {
    "groups": [
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qkv.weight",
          "blocks.0.attn_out.weight",
          "blocks.0.gate_up.weight",
          "blocks.0.down.weight",
          "blocks.1.d_write.context_key.weight",
          "blocks.1.d_write.delta_key.0.weight",
          "blocks.1.d_write.delta_key.1.weight",
          "blocks.1.d_write.value.0.weight",
          "blocks.1.d_write.value.1.weight",
          "blocks.1.d_read.query.weight",
          "blocks.1.d_read.out.weight",
          "blocks.1.qkv.weight",
          "blocks.1.attn_out.weight",
          "blocks.1.gate_up.weight",
          "blocks.1.down.weight",
          "blocks.2.d_write.context_key.weight",
          "blocks.2.d_write.delta_key.0.weight",
          "blocks.2.d_write.delta_key.1.weight",
          "blocks.2.d_write.value.0.weight",
          "blocks.2.d_write.value.1.weight",
          "blocks.2.d_read.query.weight",
          "blocks.2.d_read.out.weight",
          "blocks.2.qkv.weight",
          "blocks.2.attn_out.weight",
          "blocks.2.gate_up.weight",
          "blocks.2.down.weight",
          "blocks.3.d_write.context_key.weight",
          "blocks.3.d_write.delta_key.0.weight",
          "blocks.3.d_write.delta_key.1.weight",
          "blocks.3.d_write.value.0.weight",
          "blocks.3.d_write.value.1.weight",
          "blocks.3.d_read.query.weight",
          "blocks.3.d_read.out.weight",
          "blocks.3.qkv.weight",
          "blocks.3.attn_out.weight",
          "blocks.3.gate_up.weight",
          "blocks.3.down.weight",
          "blocks.4.d_write.context_key.weight",
          "blocks.4.d_write.delta_key.0.weight",
          "blocks.4.d_write.delta_key.1.weight",
          "blocks.4.d_write.value.0.weight",
          "blocks.4.d_write.value.1.weight",
          "blocks.4.d_read.query.weight",
          "blocks.4.d_read.out.weight",
          "blocks.4.qkv.weight",
          "blocks.4.attn_out.weight",
          "blocks.4.gate_up.weight",
          "blocks.4.down.weight",
          "blocks.5.d_write.context_key.weight",
          "blocks.5.d_write.delta_key.0.weight",
          "blocks.5.d_write.delta_key.1.weight",
          "blocks.5.d_write.value.0.weight",
          "blocks.5.d_write.value.1.weight",
          "blocks.5.d_read.query.weight",
          "blocks.5.d_read.out.weight",
          "blocks.5.qkv.weight",
          "blocks.5.attn_out.weight",
          "blocks.5.gate_up.weight",
          "blocks.5.down.weight",
          "blocks.6.d_write.context_key.weight",
          "blocks.6.d_write.delta_key.0.weight",
          "blocks.6.d_write.delta_key.1.weight",
          "blocks.6.d_write.value.0.weight",
          "blocks.6.d_write.value.1.weight",
          "blocks.6.d_read.query.weight",
          "blocks.6.d_read.out.weight",
          "blocks.6.qkv.weight",
          "blocks.6.attn_out.weight",
          "blocks.6.gate_up.weight",
          "blocks.6.down.weight",
          "blocks.7.d_write.context_key.weight",
          "blocks.7.d_write.delta_key.0.weight",
          "blocks.7.d_write.delta_key.1.weight",
          "blocks.7.d_write.value.0.weight",
          "blocks.7.d_write.value.1.weight",
          "blocks.7.d_read.query.weight",
          "blocks.7.d_read.out.weight",
          "blocks.7.qkv.weight",
          "blocks.7.attn_out.weight",
          "blocks.7.gate_up.weight",
          "blocks.7.down.weight",
          "embedding.weight",
          "head.weight"
        ],
        "label": "blocks.*.qkv.weight, blocks.*.attn_out.weight, blocks.*.gate_up.weight, blocks.*.down.weight, blocks.*.d_write.context_key.weight, blocks.*.d_write.delta_key.0.weight, blocks.*.d_write.delta_key.1.weight, blocks.*.d_write.value.0.weight, blocks.*.d_write.value.1.weight, blocks.*.d_read.query.weight, blocks.*.d_read.out.weight, embedding.weight, head.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.raw_gamma",
          "blocks.0.a_res.raw_eta",
          "blocks.0.f_res.raw_eta",
          "blocks.1.raw_gamma",
          "blocks.1.d_write.retention.proj.weight",
          "blocks.1.d_write.retention.proj.bias",
          "blocks.1.d_write.write.proj.weight",
          "blocks.1.d_write.write.proj.bias",
          "blocks.1.d_write.allocation.weight",
          "blocks.1.d_write.allocation.bias",
          "blocks.1.d_read.residual.raw_eta",
          "blocks.1.a_res.raw_eta",
          "blocks.1.f_res.raw_eta",
          "blocks.2.raw_gamma",
          "blocks.2.d_write.retention.proj.weight",
          "blocks.2.d_write.retention.proj.bias",
          "blocks.2.d_write.write.proj.weight",
          "blocks.2.d_write.write.proj.bias",
          "blocks.2.d_write.allocation.weight",
          "blocks.2.d_write.allocation.bias",
          "blocks.2.d_read.residual.raw_eta",
          "blocks.2.a_res.raw_eta",
          "blocks.2.f_res.raw_eta",
          "blocks.3.raw_gamma",
          "blocks.3.d_write.retention.proj.weight",
          "blocks.3.d_write.retention.proj.bias",
          "blocks.3.d_write.write.proj.weight",
          "blocks.3.d_write.write.proj.bias",
          "blocks.3.d_write.allocation.weight",
          "blocks.3.d_write.allocation.bias",
          "blocks.3.d_read.residual.raw_eta",
          "blocks.3.a_res.raw_eta",
          "blocks.3.f_res.raw_eta",
          "blocks.4.raw_gamma",
          "blocks.4.d_write.retention.proj.weight",
          "blocks.4.d_write.retention.proj.bias",
          "blocks.4.d_write.write.proj.weight",
          "blocks.4.d_write.write.proj.bias",
          "blocks.4.d_write.allocation.weight",
          "blocks.4.d_write.allocation.bias",
          "blocks.4.d_read.residual.raw_eta",
          "blocks.4.a_res.raw_eta",
          "blocks.4.f_res.raw_eta",
          "blocks.5.raw_gamma",
          "blocks.5.d_write.retention.proj.weight",
          "blocks.5.d_write.retention.proj.bias",
          "blocks.5.d_write.write.proj.weight",
          "blocks.5.d_write.write.proj.bias",
          "blocks.5.d_write.allocation.weight",
          "blocks.5.d_write.allocation.bias",
          "blocks.5.d_read.residual.raw_eta",
          "blocks.5.a_res.raw_eta",
          "blocks.5.f_res.raw_eta",
          "blocks.6.raw_gamma",
          "blocks.6.d_write.retention.proj.weight",
          "blocks.6.d_write.retention.proj.bias",
          "blocks.6.d_write.write.proj.weight",
          "blocks.6.d_write.write.proj.bias",
          "blocks.6.d_write.allocation.weight",
          "blocks.6.d_write.allocation.bias",
          "blocks.6.d_read.residual.raw_eta",
          "blocks.6.a_res.raw_eta",
          "blocks.6.f_res.raw_eta",
          "blocks.7.raw_gamma",
          "blocks.7.d_write.retention.proj.weight",
          "blocks.7.d_write.retention.proj.bias",
          "blocks.7.d_write.write.proj.weight",
          "blocks.7.d_write.write.proj.bias",
          "blocks.7.d_write.allocation.weight",
          "blocks.7.d_write.allocation.bias",
          "blocks.7.d_read.residual.raw_eta",
          "blocks.7.a_res.raw_eta",
          "blocks.7.f_res.raw_eta"
        ],
        "label": "blocks.*.raw_gamma, blocks.*.a_res.raw_eta, blocks.*.f_res.raw_eta, blocks.*.d_write.retention.proj.weight, blocks.*.d_write.retention.proj.bias, blocks.*.d_write.write.proj.weight, blocks.*.d_write.write.proj.bias, blocks.*.d_write.allocation.weight, blocks.*.d_write.allocation.bias, blocks.*.d_read.residual.raw_eta"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0004,
      0.00035,
      0.0003
    ],
    "schedule": {
      "lr": [
        0.0004,
        0.00035,
        0.0003
      ],
      "start_words": [
        0,
        30000000,
        60000000
      ],
      "warmup_words": 1600000
    },
    "optimizer": {
      "type": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "mean supervised-token causal NLL; matched logical-window and group masks",
    "initialization": "all156 tensors exact #298",
    "runtime_optimizations": [
      "FP32 recompute D",
      "fused compiled body",
      "padded microbatch budget8192",
      "compile cache64"
    ]
  },
  {
    "groups": [
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qkv.weight",
          "blocks.0.attn_out.weight",
          "blocks.0.gate_up.weight",
          "blocks.0.down.weight",
          "blocks.1.d_write.context_key.weight",
          "blocks.1.d_write.delta_key.0.weight",
          "blocks.1.d_write.delta_key.1.weight",
          "blocks.1.d_write.value.0.weight",
          "blocks.1.d_write.value.1.weight",
          "blocks.1.d_read.query.weight",
          "blocks.1.d_read.out.weight",
          "blocks.1.qkv.weight",
          "blocks.1.attn_out.weight",
          "blocks.1.gate_up.weight",
          "blocks.1.down.weight",
          "blocks.2.d_write.context_key.weight",
          "blocks.2.d_write.delta_key.0.weight",
          "blocks.2.d_write.delta_key.1.weight",
          "blocks.2.d_write.value.0.weight",
          "blocks.2.d_write.value.1.weight",
          "blocks.2.d_read.query.weight",
          "blocks.2.d_read.out.weight",
          "blocks.2.qkv.weight",
          "blocks.2.attn_out.weight",
          "blocks.2.gate_up.weight",
          "blocks.2.down.weight",
          "blocks.3.d_write.context_key.weight",
          "blocks.3.d_write.delta_key.0.weight",
          "blocks.3.d_write.delta_key.1.weight",
          "blocks.3.d_write.value.0.weight",
          "blocks.3.d_write.value.1.weight",
          "blocks.3.d_read.query.weight",
          "blocks.3.d_read.out.weight",
          "blocks.3.qkv.weight",
          "blocks.3.attn_out.weight",
          "blocks.3.gate_up.weight",
          "blocks.3.down.weight",
          "blocks.4.d_write.context_key.weight",
          "blocks.4.d_write.delta_key.0.weight",
          "blocks.4.d_write.delta_key.1.weight",
          "blocks.4.d_write.value.0.weight",
          "blocks.4.d_write.value.1.weight",
          "blocks.4.d_read.query.weight",
          "blocks.4.d_read.out.weight",
          "blocks.4.qkv.weight",
          "blocks.4.attn_out.weight",
          "blocks.4.gate_up.weight",
          "blocks.4.down.weight",
          "blocks.5.d_write.context_key.weight",
          "blocks.5.d_write.delta_key.0.weight",
          "blocks.5.d_write.delta_key.1.weight",
          "blocks.5.d_write.value.0.weight",
          "blocks.5.d_write.value.1.weight",
          "blocks.5.d_read.query.weight",
          "blocks.5.d_read.out.weight",
          "blocks.5.qkv.weight",
          "blocks.5.attn_out.weight",
          "blocks.5.gate_up.weight",
          "blocks.5.down.weight",
          "blocks.6.d_write.context_key.weight",
          "blocks.6.d_write.delta_key.0.weight",
          "blocks.6.d_write.delta_key.1.weight",
          "blocks.6.d_write.value.0.weight",
          "blocks.6.d_write.value.1.weight",
          "blocks.6.d_read.query.weight",
          "blocks.6.d_read.out.weight",
          "blocks.6.qkv.weight",
          "blocks.6.attn_out.weight",
          "blocks.6.gate_up.weight",
          "blocks.6.down.weight",
          "blocks.7.d_write.context_key.weight",
          "blocks.7.d_write.delta_key.0.weight",
          "blocks.7.d_write.delta_key.1.weight",
          "blocks.7.d_write.value.0.weight",
          "blocks.7.d_write.value.1.weight",
          "blocks.7.d_read.query.weight",
          "blocks.7.d_read.out.weight",
          "blocks.7.qkv.weight",
          "blocks.7.attn_out.weight",
          "blocks.7.gate_up.weight",
          "blocks.7.down.weight",
          "embedding.weight",
          "head.weight"
        ],
        "label": "blocks.*.qkv.weight, blocks.*.attn_out.weight, blocks.*.gate_up.weight, blocks.*.down.weight, blocks.*.d_write.context_key.weight, blocks.*.d_write.delta_key.0.weight, blocks.*.d_write.delta_key.1.weight, blocks.*.d_write.value.0.weight, blocks.*.d_write.value.1.weight, blocks.*.d_read.query.weight, blocks.*.d_read.out.weight, embedding.weight, head.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.raw_gamma",
          "blocks.0.a_res.raw_eta",
          "blocks.0.f_res.raw_eta",
          "blocks.1.raw_gamma",
          "blocks.1.d_write.retention.proj.weight",
          "blocks.1.d_write.retention.proj.bias",
          "blocks.1.d_write.write.proj.weight",
          "blocks.1.d_write.write.proj.bias",
          "blocks.1.d_write.allocation.weight",
          "blocks.1.d_write.allocation.bias",
          "blocks.1.d_read.residual.raw_eta",
          "blocks.1.d_read.adoption.weight",
          "blocks.1.d_read.adoption.bias",
          "blocks.1.a_res.raw_eta",
          "blocks.1.f_res.raw_eta",
          "blocks.2.raw_gamma",
          "blocks.2.d_write.retention.proj.weight",
          "blocks.2.d_write.retention.proj.bias",
          "blocks.2.d_write.write.proj.weight",
          "blocks.2.d_write.write.proj.bias",
          "blocks.2.d_write.allocation.weight",
          "blocks.2.d_write.allocation.bias",
          "blocks.2.d_read.residual.raw_eta",
          "blocks.2.d_read.adoption.weight",
          "blocks.2.d_read.adoption.bias",
          "blocks.2.a_res.raw_eta",
          "blocks.2.f_res.raw_eta",
          "blocks.3.raw_gamma",
          "blocks.3.d_write.retention.proj.weight",
          "blocks.3.d_write.retention.proj.bias",
          "blocks.3.d_write.write.proj.weight",
          "blocks.3.d_write.write.proj.bias",
          "blocks.3.d_write.allocation.weight",
          "blocks.3.d_write.allocation.bias",
          "blocks.3.d_read.residual.raw_eta",
          "blocks.3.d_read.adoption.weight",
          "blocks.3.d_read.adoption.bias",
          "blocks.3.a_res.raw_eta",
          "blocks.3.f_res.raw_eta",
          "blocks.4.raw_gamma",
          "blocks.4.d_write.retention.proj.weight",
          "blocks.4.d_write.retention.proj.bias",
          "blocks.4.d_write.write.proj.weight",
          "blocks.4.d_write.write.proj.bias",
          "blocks.4.d_write.allocation.weight",
          "blocks.4.d_write.allocation.bias",
          "blocks.4.d_read.residual.raw_eta",
          "blocks.4.d_read.adoption.weight",
          "blocks.4.d_read.adoption.bias",
          "blocks.4.a_res.raw_eta",
          "blocks.4.f_res.raw_eta",
          "blocks.5.raw_gamma",
          "blocks.5.d_write.retention.proj.weight",
          "blocks.5.d_write.retention.proj.bias",
          "blocks.5.d_write.write.proj.weight",
          "blocks.5.d_write.write.proj.bias",
          "blocks.5.d_write.allocation.weight",
          "blocks.5.d_write.allocation.bias",
          "blocks.5.d_read.residual.raw_eta",
          "blocks.5.d_read.adoption.weight",
          "blocks.5.d_read.adoption.bias",
          "blocks.5.a_res.raw_eta",
          "blocks.5.f_res.raw_eta",
          "blocks.6.raw_gamma",
          "blocks.6.d_write.retention.proj.weight",
          "blocks.6.d_write.retention.proj.bias",
          "blocks.6.d_write.write.proj.weight",
          "blocks.6.d_write.write.proj.bias",
          "blocks.6.d_write.allocation.weight",
          "blocks.6.d_write.allocation.bias",
          "blocks.6.d_read.residual.raw_eta",
          "blocks.6.d_read.adoption.weight",
          "blocks.6.d_read.adoption.bias",
          "blocks.6.a_res.raw_eta",
          "blocks.6.f_res.raw_eta",
          "blocks.7.raw_gamma",
          "blocks.7.d_write.retention.proj.weight",
          "blocks.7.d_write.retention.proj.bias",
          "blocks.7.d_write.write.proj.weight",
          "blocks.7.d_write.write.proj.bias",
          "blocks.7.d_write.allocation.weight",
          "blocks.7.d_write.allocation.bias",
          "blocks.7.d_read.residual.raw_eta",
          "blocks.7.d_read.adoption.weight",
          "blocks.7.d_read.adoption.bias",
          "blocks.7.a_res.raw_eta",
          "blocks.7.f_res.raw_eta"
        ],
        "label": "blocks.*.raw_gamma, blocks.*.a_res.raw_eta, blocks.*.f_res.raw_eta, blocks.*.d_write.retention.proj.weight, blocks.*.d_write.retention.proj.bias, blocks.*.d_write.write.proj.weight, blocks.*.d_write.write.proj.bias, blocks.*.d_write.allocation.weight, blocks.*.d_write.allocation.bias, blocks.*.d_read.residual.raw_eta, blocks.*.d_read.adoption.weight, blocks.*.d_read.adoption.bias"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0004,
      0.00035,
      0.0003
    ],
    "schedule": {
      "lr": [
        0.0004,
        0.00035,
        0.0003
      ],
      "start_words": [
        0,
        30000000,
        60000000
      ],
      "warmup_words": 1600000
    },
    "optimizer": {
      "type": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "mean supervised-token causal NLL; matched logical-window and group masks",
    "initialization": "common156 tensors and post-init RNG exact #299; new14 gate tensors zeros",
    "runtime_optimizations": [
      "FP32 recompute D",
      "fused compiled body",
      "padded microbatch budget8192",
      "compile cache64"
    ]
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "Every per-name numerical coefficient frozen to parent320, including FFNdown0.8; no further width-derived LR/WD adjustment"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent320 numeric std/multipliers, per-name seed, unchanged-shaped tensors and zero-word training RNG; new shapes generated fresh; native MoS compress std anchored to320",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Restore actual40M model, optimizer, RNG and next unread data cursor. No fresh initialization or warmup reset.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Restore actual40M model, optimizer, RNG and next unread data cursor. No fresh initialization or warmup reset.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent initialization law under seed731946281; no restored old RNG; FFN1121 preserves new-seed parent coordinates and separate extra dropout.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged",
      "main_matrix_lr_schedule": {
        "values": [
          0.0007,
          0.0006,
          0.0005
        ],
        "boundaries": [
          30000000,
          60000000
        ],
        "selection": "role matrices except embed.weight; Candidate Reader and scalar/bias/Norm parameters excluded",
        "semantics": "absolute effective LR after warmup; applied before optimizer.step"
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent initialization law under seed731946281; no restored old RNG; FFN1121 preserves new-seed parent coordinates and separate extra dropout.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      }
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent initialization law under seed731946281; no restored old RNG; FFN1121 preserves new-seed parent coordinates and separate extra dropout.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Parent decoder implementation retained; MoS3 CUDA graph and runtime verification pending"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.0.candidate_read.query.weight",
          "blocks.0.candidate_read.readout.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.1.candidate_read.query.weight",
          "blocks.1.candidate_read.readout.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.2.candidate_read.query.weight",
          "blocks.2.candidate_read.readout.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.3.candidate_read.query.weight",
          "blocks.3.candidate_read.readout.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.4.candidate_read.query.weight",
          "blocks.4.candidate_read.readout.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.5.candidate_read.query.weight",
          "blocks.5.candidate_read.readout.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.6.candidate_read.query.weight",
          "blocks.6.candidate_read.readout.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.7.candidate_read.query.weight",
          "blocks.7.candidate_read.readout.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "blocks.8.candidate_read.query.weight",
          "blocks.8.candidate_read.readout.weight",
          "lm_head.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.candidate_read.query.weight, blocks.*.candidate_read.readout.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "Exact parent320 per-parameter lr_scale and weight_decay table; current-width initialization remains fan-in based and is independent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Single final-state next-token log-softmax NLL; no MoS and no auxiliary loss; gradient through both parallel passes.",
    "initialization": "All shared same-shape initial tensors use successful parent336FFN1232 factory seed/fan-in formula; Router32 retained; learned bias/variance/RoPE initial values retained. New candidate query named-seeded; readout zero.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.0.candidate_read.query.weight",
          "blocks.0.candidate_read.readout.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.1.candidate_read.query.weight",
          "blocks.1.candidate_read.readout.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.2.candidate_read.query.weight",
          "blocks.2.candidate_read.readout.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.3.candidate_read.query.weight",
          "blocks.3.candidate_read.readout.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.4.candidate_read.query.weight",
          "blocks.4.candidate_read.readout.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.5.candidate_read.query.weight",
          "blocks.5.candidate_read.readout.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.6.candidate_read.query.weight",
          "blocks.6.candidate_read.readout.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.7.candidate_read.query.weight",
          "blocks.7.candidate_read.readout.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "blocks.8.candidate_read.query.weight",
          "blocks.8.candidate_read.readout.weight",
          "lm_head.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.8.attn_gate.received.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.candidate_read.query.weight, blocks.*.candidate_read.readout.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, blocks.*.attn_gate.received.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "Exact parent320 per-parameter lr_scale and weight_decay table; current-width initialization remains fan-in based and is independent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Single final-state next-token log-softmax NLL; no MoS and no auxiliary loss; gradient through both parallel passes.",
    "initialization": "All shared same-shape initial tensors use successful parent336FFN1232 factory seed/fan-in formula; Router32 retained; learned bias/variance/RoPE initial values retained. New candidate query named-seeded; readout zero. Added received matrices are zero; common initial tensors and constructor RNG are preserved.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias",
          "blocks.0.ffn.bypass",
          "blocks.1.ffn.bypass",
          "blocks.2.ffn.bypass",
          "blocks.3.ffn.bypass",
          "blocks.4.ffn.bypass",
          "blocks.5.ffn.bypass",
          "blocks.6.ffn.bypass",
          "blocks.7.ffn.bypass",
          "blocks.8.ffn.bypass"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias, blocks.*.ffn.bypass"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.0.candidate_read.query.weight",
          "blocks.0.candidate_read.readout.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.1.candidate_read.query.weight",
          "blocks.1.candidate_read.readout.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.2.candidate_read.query.weight",
          "blocks.2.candidate_read.readout.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.3.candidate_read.query.weight",
          "blocks.3.candidate_read.readout.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.4.candidate_read.query.weight",
          "blocks.4.candidate_read.readout.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.5.candidate_read.query.weight",
          "blocks.5.candidate_read.readout.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.6.candidate_read.query.weight",
          "blocks.6.candidate_read.readout.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.7.candidate_read.query.weight",
          "blocks.7.candidate_read.readout.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "blocks.8.candidate_read.query.weight",
          "blocks.8.candidate_read.readout.weight",
          "lm_head.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.8.attn_gate.received.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.candidate_read.query.weight, blocks.*.candidate_read.readout.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, blocks.*.attn_gate.received.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "Exact parent320 per-parameter lr_scale and weight_decay table; current-width initialization remains fan-in based and is independent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Single final-state next-token log-softmax NLL; no MoS and no auxiliary loss; gradient through both parallel passes.",
    "initialization": "Fresh named-seeded model, all common tensors equal current received-gate parent; addedFFN bypass vectors zero, no RNG consumption.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.0.candidate_read.query.weight",
          "blocks.0.candidate_read.readout.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.1.candidate_read.query.weight",
          "blocks.1.candidate_read.readout.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.2.candidate_read.query.weight",
          "blocks.2.candidate_read.readout.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.3.candidate_read.query.weight",
          "blocks.3.candidate_read.readout.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.4.candidate_read.query.weight",
          "blocks.4.candidate_read.readout.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.5.candidate_read.query.weight",
          "blocks.5.candidate_read.readout.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.6.candidate_read.query.weight",
          "blocks.6.candidate_read.readout.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.7.candidate_read.query.weight",
          "blocks.7.candidate_read.readout.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "blocks.8.candidate_read.query.weight",
          "blocks.8.candidate_read.readout.weight",
          "lm_head.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.8.attn_gate.received.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.candidate_read.query.weight, blocks.*.candidate_read.readout.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, blocks.*.attn_gate.received.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "Exact parent320 per-parameter lr_scale and weight_decay table; current-width initialization remains fan-in based and is independent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final MoS3 mixed-probability next-token NLL; no auxiliary loss; gradient through both parallel passes and all3 branches.",
    "initialization": "All shared same-shape initial tensors use successful parent336FFN1232 factory seed/fan-in formula; Router32 retained; learned bias/variance/RoPE initial values retained. New candidate query named-seeded; readout zero. Added received matrices are zero; common initial tensors and constructor RNG are preserved. MoS3 additions use donor initialized tensors and effective updates.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw",
          "candidate_variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw, candidate_variance_raw"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.0.candidate_read.query.weight",
          "blocks.0.candidate_read.readout.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.1.candidate_read.query.weight",
          "blocks.1.candidate_read.readout.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.2.candidate_read.query.weight",
          "blocks.2.candidate_read.readout.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.3.candidate_read.query.weight",
          "blocks.3.candidate_read.readout.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.4.candidate_read.query.weight",
          "blocks.4.candidate_read.readout.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.5.candidate_read.query.weight",
          "blocks.5.candidate_read.readout.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.6.candidate_read.query.weight",
          "blocks.6.candidate_read.readout.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.7.candidate_read.query.weight",
          "blocks.7.candidate_read.readout.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "blocks.8.candidate_read.query.weight",
          "blocks.8.candidate_read.readout.weight",
          "lm_head.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.8.attn_gate.received.weight",
          "multi_output.compress1.weight",
          "multi_output.compress2.weight",
          "multi_output.compress3.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.candidate_read.query.weight, blocks.*.candidate_read.readout.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, blocks.*.attn_gate.received.weight, multi_output.compress1.weight, multi_output.compress2.weight, multi_output.compress3.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "Exact parent320 per-parameter lr_scale and weight_decay table; current-width initialization remains fan-in based and is independent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final MoS3 mixed-probability next-token NLL; no auxiliary loss; gradient through both parallel passes and all3 branches.",
    "initialization": "All shared same-shape initial tensors use successful parent336FFN1232 factory seed/fan-in formula; Router32 retained; learned bias/variance/RoPE initial values retained. New candidate query named-seeded; readout zero. Added received matrices are zero; common initial tensors and constructor RNG are preserved. New MoS3 parameters match donor336 tensors; existing174 tensors and RNG remain identical. Reader48 donor parameters added; preserve current MoS3 parent181parameter initial tensors and training RNG.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw",
          "candidate_variance_raw"
        ],
        "label": "feedback_threshold_raw, blocks.*.ffn.variance_raw, candidate_variance_raw"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.0.candidate_read.query.weight",
          "blocks.0.candidate_read.readout.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.1.candidate_read.query.weight",
          "blocks.1.candidate_read.readout.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.2.candidate_read.query.weight",
          "blocks.2.candidate_read.readout.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.3.candidate_read.query.weight",
          "blocks.3.candidate_read.readout.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.4.candidate_read.query.weight",
          "blocks.4.candidate_read.readout.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.5.candidate_read.query.weight",
          "blocks.5.candidate_read.readout.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.6.candidate_read.query.weight",
          "blocks.6.candidate_read.readout.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.7.candidate_read.query.weight",
          "blocks.7.candidate_read.readout.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "blocks.8.candidate_read.query.weight",
          "blocks.8.candidate_read.readout.weight",
          "lm_head.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.8.attn_gate.received.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.candidate_read.query.weight, blocks.*.candidate_read.readout.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, blocks.*.attn_gate.received.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.0003206201550387597,
          0.0002805426356589147,
          0.00024046511627906974
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "Exact parent320 per-parameter lr_scale and weight_decay table; current-width initialization remains fan-in based and is independent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Single final-state next-token log-softmax NLL; no MoS and no auxiliary loss; gradient through both parallel passes.",
    "initialization": "All shared same-shape initial tensors use successful parent336FFN1232 factory seed/fan-in formula; Router32 retained; learned bias/variance/RoPE initial values retained. New candidate query named-seeded; readout zero. Added received matrices are zero; common initial tensors and constructor RNG are preserved.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw"
        ],
        "label": "feedback_beta, temperature_raw"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw"
        ],
        "label": "feedback_threshold_raw"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.00064,
          0.0005600000000000001,
          0.00047999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.0.candidate_read.query.weight",
          "blocks.0.candidate_read.readout.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.1.candidate_read.query.weight",
          "blocks.1.candidate_read.readout.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.2.candidate_read.query.weight",
          "blocks.2.candidate_read.readout.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.3.candidate_read.query.weight",
          "blocks.3.candidate_read.readout.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.4.candidate_read.query.weight",
          "blocks.4.candidate_read.readout.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.5.candidate_read.query.weight",
          "blocks.5.candidate_read.readout.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.6.candidate_read.query.weight",
          "blocks.6.candidate_read.readout.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.7.candidate_read.query.weight",
          "blocks.7.candidate_read.readout.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "blocks.8.candidate_read.query.weight",
          "blocks.8.candidate_read.readout.weight",
          "lm_head.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.candidate_read.query.weight, blocks.*.candidate_read.readout.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight"
      },
      {
        "lr": [
          0.0006404984423676013,
          0.0005604361370716511,
          0.00048037383177570086
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        },
        "ffn_bypass": {
          "lr_scale": 0.3,
          "weight_decay": 0.0
        }
      },
      "width_lr_composition": "Exact parent320 per-parameter lr_scale and weight_decay table; current-width initialization remains fan-in based and is independent."
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Parent named-tensor and fan-in formula at unchanged dimensions; Router32to16 shape changes; query named-seeded; candidate readout zero. Common effective LR/WD unchanged.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 TF32off prefix graph; measured parity, throughput and full cache capacity before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0005333333333333334,
          0.00046666666666666666,
          0.00039999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.mixing.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.mixing.weight"
      },
      {
        "lr": [
          0.0005340259740259741,
          0.00046727272727272725,
          0.0004005194805194805
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.00026752910737386806,
          0.00023408796895213455,
          0.00020064683053040103
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:sc12367b59cf1b9",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Fresh384 matrix initialization, reference256 fan-in scaling; preserve MoS facet identity halves, native compression scaling and zero mixtures/gates. Copy same-shape initial parameters and training RNG from existing untrained256 MoS2 fixture in RAM; empty Adam. No initial or recovery checkpoint.",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Prefix CUDA graph, FP32 TF32off; real-prompt throughput and parity checked before launch"
    }
  },
  {
    "groups": [
      {
        "lr": [
          8e-05,
          7.000000000000001e-05,
          5.9999999999999995e-05
        ],
        "wd": 0.0,
        "names": [
          "rope_log_scale"
        ],
        "label": "rope_log_scale"
      },
      {
        "lr": [
          0.00048,
          0.00041999999999999996,
          0.00035999999999999997
        ],
        "wd": 0.0,
        "names": [
          "feedback_threshold_raw",
          "candidate_variance_raw",
          "blocks.0.ffn.variance_raw",
          "blocks.1.ffn.variance_raw",
          "blocks.2.ffn.variance_raw",
          "blocks.3.ffn.variance_raw",
          "blocks.4.ffn.variance_raw",
          "blocks.5.ffn.variance_raw",
          "blocks.6.ffn.variance_raw",
          "blocks.7.ffn.variance_raw",
          "blocks.8.ffn.variance_raw"
        ],
        "label": "feedback_threshold_raw, candidate_variance_raw, blocks.*.ffn.variance_raw"
      },
      {
        "lr": [
          0.00024,
          0.00020999999999999998,
          0.00017999999999999998
        ],
        "wd": 0.0,
        "names": [
          "feedback_beta",
          "temperature_raw",
          "blocks.0.attn_gate.proj.bias",
          "blocks.1.attn_gate.proj.bias",
          "blocks.2.attn_gate.proj.bias",
          "blocks.3.attn_gate.proj.bias",
          "blocks.4.attn_gate.proj.bias",
          "blocks.5.attn_gate.proj.bias",
          "blocks.6.attn_gate.proj.bias",
          "blocks.7.attn_gate.proj.bias",
          "blocks.8.attn_gate.proj.bias"
        ],
        "label": "feedback_beta, temperature_raw, blocks.*.attn_gate.proj.bias"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.1,
        "names": [
          "embed.weight"
        ],
        "label": "embed.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.attention_log_gain",
          "blocks.0.an.weight",
          "blocks.0.fn.weight",
          "blocks.1.attention_log_gain",
          "blocks.1.an.weight",
          "blocks.1.fn.weight",
          "blocks.2.attention_log_gain",
          "blocks.2.an.weight",
          "blocks.2.fn.weight",
          "blocks.3.attention_log_gain",
          "blocks.3.an.weight",
          "blocks.3.fn.weight",
          "blocks.4.attention_log_gain",
          "blocks.4.an.weight",
          "blocks.4.fn.weight",
          "blocks.5.attention_log_gain",
          "blocks.5.an.weight",
          "blocks.5.fn.weight",
          "blocks.6.attention_log_gain",
          "blocks.6.an.weight",
          "blocks.6.fn.weight",
          "blocks.7.attention_log_gain",
          "blocks.7.an.weight",
          "blocks.7.fn.weight",
          "blocks.8.attention_log_gain",
          "blocks.8.an.weight",
          "blocks.8.fn.weight",
          "norm.weight"
        ],
        "label": "blocks.*.attention_log_gain, blocks.*.an.weight, blocks.*.fn.weight, norm.weight"
      },
      {
        "lr": [
          0.0005333333333333334,
          0.00046666666666666666,
          0.00039999999999999996
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.qk.weight",
          "blocks.0.v.weight",
          "blocks.0.attn_gate.proj.weight",
          "blocks.0.attn_gate.received.weight",
          "blocks.0.out.weight",
          "blocks.0.ffn.base.weight",
          "blocks.0.ffn.down.weight",
          "blocks.1.qk.weight",
          "blocks.1.v.weight",
          "blocks.1.attn_gate.proj.weight",
          "blocks.1.attn_gate.received.weight",
          "blocks.1.out.weight",
          "blocks.1.ffn.base.weight",
          "blocks.1.ffn.down.weight",
          "blocks.1.ffn.reader.aux1.weight",
          "blocks.1.ffn.reader.aux2.weight",
          "blocks.1.ffn.reader.router.query.weight",
          "blocks.1.ffn.reader.router.key.weight",
          "blocks.2.qk.weight",
          "blocks.2.v.weight",
          "blocks.2.attn_gate.proj.weight",
          "blocks.2.attn_gate.received.weight",
          "blocks.2.out.weight",
          "blocks.2.ffn.base.weight",
          "blocks.2.ffn.down.weight",
          "blocks.2.ffn.reader.aux1.weight",
          "blocks.2.ffn.reader.aux2.weight",
          "blocks.2.ffn.reader.router.query.weight",
          "blocks.2.ffn.reader.router.key.weight",
          "blocks.3.qk.weight",
          "blocks.3.v.weight",
          "blocks.3.attn_gate.proj.weight",
          "blocks.3.attn_gate.received.weight",
          "blocks.3.out.weight",
          "blocks.3.ffn.base.weight",
          "blocks.3.ffn.down.weight",
          "blocks.3.ffn.reader.aux1.weight",
          "blocks.3.ffn.reader.aux2.weight",
          "blocks.3.ffn.reader.router.query.weight",
          "blocks.3.ffn.reader.router.key.weight",
          "blocks.4.qk.weight",
          "blocks.4.v.weight",
          "blocks.4.attn_gate.proj.weight",
          "blocks.4.attn_gate.received.weight",
          "blocks.4.out.weight",
          "blocks.4.ffn.base.weight",
          "blocks.4.ffn.down.weight",
          "blocks.4.ffn.reader.aux1.weight",
          "blocks.4.ffn.reader.aux2.weight",
          "blocks.4.ffn.reader.router.query.weight",
          "blocks.4.ffn.reader.router.key.weight",
          "blocks.5.qk.weight",
          "blocks.5.v.weight",
          "blocks.5.attn_gate.proj.weight",
          "blocks.5.attn_gate.received.weight",
          "blocks.5.out.weight",
          "blocks.5.ffn.base.weight",
          "blocks.5.ffn.down.weight",
          "blocks.5.ffn.reader.aux1.weight",
          "blocks.5.ffn.reader.aux2.weight",
          "blocks.5.ffn.reader.router.query.weight",
          "blocks.5.ffn.reader.router.key.weight",
          "blocks.6.qk.weight",
          "blocks.6.v.weight",
          "blocks.6.attn_gate.proj.weight",
          "blocks.6.attn_gate.received.weight",
          "blocks.6.out.weight",
          "blocks.6.ffn.base.weight",
          "blocks.6.ffn.down.weight",
          "blocks.6.ffn.reader.aux1.weight",
          "blocks.6.ffn.reader.aux2.weight",
          "blocks.6.ffn.reader.router.query.weight",
          "blocks.6.ffn.reader.router.key.weight",
          "blocks.7.qk.weight",
          "blocks.7.v.weight",
          "blocks.7.attn_gate.proj.weight",
          "blocks.7.attn_gate.received.weight",
          "blocks.7.out.weight",
          "blocks.7.ffn.base.weight",
          "blocks.7.ffn.down.weight",
          "blocks.7.ffn.reader.aux1.weight",
          "blocks.7.ffn.reader.aux2.weight",
          "blocks.7.ffn.reader.router.query.weight",
          "blocks.7.ffn.reader.router.key.weight",
          "blocks.8.qk.weight",
          "blocks.8.v.weight",
          "blocks.8.attn_gate.proj.weight",
          "blocks.8.attn_gate.received.weight",
          "blocks.8.out.weight",
          "blocks.8.ffn.base.weight",
          "blocks.8.ffn.down.weight",
          "blocks.8.ffn.reader.aux1.weight",
          "blocks.8.ffn.reader.aux2.weight",
          "blocks.8.ffn.reader.router.query.weight",
          "blocks.8.ffn.reader.router.key.weight",
          "lm_head.weight",
          "multi_output.compress.weight",
          "multi_output.facet1.weight",
          "multi_output.facet2.weight",
          "multi_output.facet3.weight",
          "multi_output.mixing.weight",
          "multi_output.compress_short.weight"
        ],
        "label": "blocks.*.qk.weight, blocks.*.v.weight, blocks.*.attn_gate.proj.weight, blocks.*.attn_gate.received.weight, blocks.*.out.weight, blocks.*.ffn.base.weight, blocks.*.ffn.down.weight, blocks.*.ffn.reader.aux1.weight, blocks.*.ffn.reader.aux2.weight, blocks.*.ffn.reader.router.query.weight, blocks.*.ffn.reader.router.key.weight, lm_head.weight, multi_output.compress.weight, multi_output.facet1.weight, multi_output.facet2.weight, multi_output.facet3.weight, multi_output.mixing.weight, multi_output.compress_short.weight"
      },
      {
        "lr": [
          0.0005340259740259741,
          0.00046727272727272725,
          0.0004005194805194805
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.ffn.spread_input.weight",
          "blocks.1.ffn.spread_input.weight",
          "blocks.2.ffn.spread_input.weight",
          "blocks.3.ffn.spread_input.weight",
          "blocks.4.ffn.spread_input.weight",
          "blocks.5.ffn.spread_input.weight",
          "blocks.6.ffn.spread_input.weight",
          "blocks.7.ffn.spread_input.weight",
          "blocks.8.ffn.spread_input.weight"
        ],
        "label": "blocks.*.ffn.spread_input.weight"
      },
      {
        "lr": [
          0.00026752910737386806,
          0.00023408796895213455,
          0.00020064683053040103
        ],
        "wd": 0.1,
        "names": [
          "blocks.0.candidate_reader.up.weight",
          "blocks.1.candidate_reader.up.weight",
          "blocks.2.candidate_reader.up.weight",
          "blocks.3.candidate_reader.up.weight",
          "blocks.4.candidate_reader.up.weight",
          "blocks.5.candidate_reader.up.weight",
          "blocks.6.candidate_reader.up.weight",
          "blocks.7.candidate_reader.up.weight",
          "blocks.8.candidate_reader.up.weight"
        ],
        "label": "blocks.*.candidate_reader.up.weight"
      },
      {
        "lr": [
          0.0008,
          0.0007,
          0.0006
        ],
        "wd": 0.05,
        "names": [
          "blocks.0.candidate_reader.down.weight",
          "blocks.1.candidate_reader.down.weight",
          "blocks.2.candidate_reader.down.weight",
          "blocks.3.candidate_reader.down.weight",
          "blocks.4.candidate_reader.down.weight",
          "blocks.5.candidate_reader.down.weight",
          "blocks.6.candidate_reader.down.weight",
          "blocks.7.candidate_reader.down.weight",
          "blocks.8.candidate_reader.down.weight"
        ],
        "label": "blocks.*.candidate_reader.down.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "blocks.0.candidate_reader.gate.weight",
          "blocks.0.candidate_reader.gate.bias",
          "blocks.1.candidate_reader.gate.weight",
          "blocks.1.candidate_reader.gate.bias",
          "blocks.2.candidate_reader.gate.weight",
          "blocks.2.candidate_reader.gate.bias",
          "blocks.3.candidate_reader.gate.weight",
          "blocks.3.candidate_reader.gate.bias",
          "blocks.4.candidate_reader.gate.weight",
          "blocks.4.candidate_reader.gate.bias",
          "blocks.5.candidate_reader.gate.weight",
          "blocks.5.candidate_reader.gate.bias",
          "blocks.6.candidate_reader.gate.weight",
          "blocks.6.candidate_reader.gate.bias",
          "blocks.7.candidate_reader.gate.weight",
          "blocks.7.candidate_reader.gate.bias",
          "blocks.8.candidate_reader.gate.weight",
          "blocks.8.candidate_reader.gate.bias"
        ],
        "label": "blocks.*.candidate_reader.gate.weight, blocks.*.candidate_reader.gate.bias"
      }
    ],
    "basis": "source:s4a461feac3c283",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true,
      "attention_head_gain_lr_ratio": 1.0,
      "role_overrides": {
        "alpha": {
          "lr_scale": 0.3
        },
        "temperature": {
          "lr_scale": 0.3
        },
        "threshold": {
          "lr_scale": 0.6
        },
        "rope": {
          "lr_scale": 0.1
        },
        "candidate_variance": {
          "lr_scale": 0.6
        },
        "ffn_variance": {
          "lr_scale": 0.6
        },
        "attention_gate_bias": {
          "lr_scale": 0.3
        },
        "head_gain": {
          "lr_scale": 1.0
        },
        "reader_up": {
          "lr_scale": 0.5,
          "weight_decay": 0.1
        },
        "reader_down": {
          "lr_scale": 1.0,
          "weight_decay": 0.05
        },
        "reader_gate_weight": {
          "lr_scale": 0.5,
          "weight_decay": 0.0
        },
        "reader_gate_bias": {
          "lr_scale": 0.5
        }
      },
      "width_lr_composition": "parent role multiplier times per-matrix reference_fan_in/current_fan_in; embeddings, vectors and scalar knobs unchanged"
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final mixed log probability NLL only; no auxiliary loss",
    "initialization": "Preserve384 parent fresh0M backbone, original pair compressor and facets; new h9 compressor uses separate deterministic RNG; no new model serialization",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "Inherited scoring implementation; model-specific GPU validation pending"
    }
  },
  {
    "groups": [
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.1,
        "names": [
          "c_write.key_emb.weight",
          "c_write.key_context.weight",
          "c_write.value_emb.weight",
          "blocks.0.qkv.weight",
          "blocks.0.attn_out.weight",
          "blocks.0.c_read.query.weight",
          "blocks.0.c_read.out.weight",
          "blocks.0.gate_up.weight",
          "blocks.0.down.weight",
          "blocks.1.d_write.context_key.weight",
          "blocks.1.d_write.delta_key.0.weight",
          "blocks.1.d_write.delta_key.1.weight",
          "blocks.1.d_write.value.0.weight",
          "blocks.1.d_write.value.1.weight",
          "blocks.1.d_read.query.weight",
          "blocks.1.d_read.out.weight",
          "blocks.1.qkv.weight",
          "blocks.1.attn_out.weight",
          "blocks.1.c_read.query.weight",
          "blocks.1.c_read.out.weight",
          "blocks.1.gate_up.weight",
          "blocks.1.down.weight",
          "blocks.2.d_write.context_key.weight",
          "blocks.2.d_write.delta_key.0.weight",
          "blocks.2.d_write.delta_key.1.weight",
          "blocks.2.d_write.value.0.weight",
          "blocks.2.d_write.value.1.weight",
          "blocks.2.d_read.query.weight",
          "blocks.2.d_read.out.weight",
          "blocks.2.qkv.weight",
          "blocks.2.attn_out.weight",
          "blocks.2.c_read.query.weight",
          "blocks.2.c_read.out.weight",
          "blocks.2.gate_up.weight",
          "blocks.2.down.weight",
          "blocks.3.d_write.context_key.weight",
          "blocks.3.d_write.delta_key.0.weight",
          "blocks.3.d_write.delta_key.1.weight",
          "blocks.3.d_write.value.0.weight",
          "blocks.3.d_write.value.1.weight",
          "blocks.3.d_read.query.weight",
          "blocks.3.d_read.out.weight",
          "blocks.3.qkv.weight",
          "blocks.3.attn_out.weight",
          "blocks.3.c_read.query.weight",
          "blocks.3.c_read.out.weight",
          "blocks.3.gate_up.weight",
          "blocks.3.down.weight",
          "blocks.4.d_write.context_key.weight",
          "blocks.4.d_write.delta_key.0.weight",
          "blocks.4.d_write.delta_key.1.weight",
          "blocks.4.d_write.value.0.weight",
          "blocks.4.d_write.value.1.weight",
          "blocks.4.d_read.query.weight",
          "blocks.4.d_read.out.weight",
          "blocks.4.qkv.weight",
          "blocks.4.attn_out.weight",
          "blocks.4.c_read.query.weight",
          "blocks.4.c_read.out.weight",
          "blocks.4.gate_up.weight",
          "blocks.4.down.weight",
          "blocks.5.d_write.context_key.weight",
          "blocks.5.d_write.delta_key.0.weight",
          "blocks.5.d_write.delta_key.1.weight",
          "blocks.5.d_write.value.0.weight",
          "blocks.5.d_write.value.1.weight",
          "blocks.5.d_read.query.weight",
          "blocks.5.d_read.out.weight",
          "blocks.5.qkv.weight",
          "blocks.5.attn_out.weight",
          "blocks.5.c_read.query.weight",
          "blocks.5.c_read.out.weight",
          "blocks.5.gate_up.weight",
          "blocks.5.down.weight",
          "blocks.6.d_write.context_key.weight",
          "blocks.6.d_write.delta_key.0.weight",
          "blocks.6.d_write.delta_key.1.weight",
          "blocks.6.d_write.value.0.weight",
          "blocks.6.d_write.value.1.weight",
          "blocks.6.d_read.query.weight",
          "blocks.6.d_read.out.weight",
          "blocks.6.qkv.weight",
          "blocks.6.attn_out.weight",
          "blocks.6.c_read.query.weight",
          "blocks.6.c_read.out.weight",
          "blocks.6.gate_up.weight",
          "blocks.6.down.weight",
          "blocks.7.d_write.context_key.weight",
          "blocks.7.d_write.delta_key.0.weight",
          "blocks.7.d_write.delta_key.1.weight",
          "blocks.7.d_write.value.0.weight",
          "blocks.7.d_write.value.1.weight",
          "blocks.7.d_read.query.weight",
          "blocks.7.d_read.out.weight",
          "blocks.7.qkv.weight",
          "blocks.7.attn_out.weight",
          "blocks.7.c_read.query.weight",
          "blocks.7.c_read.out.weight",
          "blocks.7.gate_up.weight",
          "blocks.7.down.weight",
          "embedding.weight",
          "head.weight"
        ],
        "label": "c_write.key_emb.weight, c_write.key_context.weight, c_write.value_emb.weight, blocks.*.qkv.weight, blocks.*.attn_out.weight, blocks.*.c_read.query.weight, blocks.*.c_read.out.weight, blocks.*.gate_up.weight, blocks.*.down.weight, blocks.*.d_write.context_key.weight, blocks.*.d_write.delta_key.0.weight, blocks.*.d_write.delta_key.1.weight, blocks.*.d_write.value.0.weight, blocks.*.d_write.value.1.weight, blocks.*.d_read.query.weight, blocks.*.d_read.out.weight, embedding.weight, head.weight"
      },
      {
        "lr": [
          0.0004,
          0.00035,
          0.0003
        ],
        "wd": 0.0,
        "names": [
          "c_write.retention.proj.weight",
          "c_write.retention.proj.bias",
          "c_write.write.proj.weight",
          "c_write.write.proj.bias",
          "c_write.threshold.weight",
          "c_write.threshold.bias",
          "blocks.0.raw_gamma",
          "blocks.0.a_res.raw_eta",
          "blocks.0.c_read.residual.raw_eta",
          "blocks.0.f_res.raw_eta",
          "blocks.1.raw_gamma",
          "blocks.1.d_write.retention.proj.weight",
          "blocks.1.d_write.retention.proj.bias",
          "blocks.1.d_write.write.proj.weight",
          "blocks.1.d_write.write.proj.bias",
          "blocks.1.d_write.allocation.weight",
          "blocks.1.d_write.allocation.bias",
          "blocks.1.d_read.residual.raw_eta",
          "blocks.1.a_res.raw_eta",
          "blocks.1.c_read.residual.raw_eta",
          "blocks.1.f_res.raw_eta",
          "blocks.2.raw_gamma",
          "blocks.2.d_write.retention.proj.weight",
          "blocks.2.d_write.retention.proj.bias",
          "blocks.2.d_write.write.proj.weight",
          "blocks.2.d_write.write.proj.bias",
          "blocks.2.d_write.allocation.weight",
          "blocks.2.d_write.allocation.bias",
          "blocks.2.d_read.residual.raw_eta",
          "blocks.2.a_res.raw_eta",
          "blocks.2.c_read.residual.raw_eta",
          "blocks.2.f_res.raw_eta",
          "blocks.3.raw_gamma",
          "blocks.3.d_write.retention.proj.weight",
          "blocks.3.d_write.retention.proj.bias",
          "blocks.3.d_write.write.proj.weight",
          "blocks.3.d_write.write.proj.bias",
          "blocks.3.d_write.allocation.weight",
          "blocks.3.d_write.allocation.bias",
          "blocks.3.d_read.residual.raw_eta",
          "blocks.3.a_res.raw_eta",
          "blocks.3.c_read.residual.raw_eta",
          "blocks.3.f_res.raw_eta",
          "blocks.4.raw_gamma",
          "blocks.4.d_write.retention.proj.weight",
          "blocks.4.d_write.retention.proj.bias",
          "blocks.4.d_write.write.proj.weight",
          "blocks.4.d_write.write.proj.bias",
          "blocks.4.d_write.allocation.weight",
          "blocks.4.d_write.allocation.bias",
          "blocks.4.d_read.residual.raw_eta",
          "blocks.4.a_res.raw_eta",
          "blocks.4.c_read.residual.raw_eta",
          "blocks.4.f_res.raw_eta",
          "blocks.5.raw_gamma",
          "blocks.5.d_write.retention.proj.weight",
          "blocks.5.d_write.retention.proj.bias",
          "blocks.5.d_write.write.proj.weight",
          "blocks.5.d_write.write.proj.bias",
          "blocks.5.d_write.allocation.weight",
          "blocks.5.d_write.allocation.bias",
          "blocks.5.d_read.residual.raw_eta",
          "blocks.5.a_res.raw_eta",
          "blocks.5.c_read.residual.raw_eta",
          "blocks.5.f_res.raw_eta",
          "blocks.6.raw_gamma",
          "blocks.6.d_write.retention.proj.weight",
          "blocks.6.d_write.retention.proj.bias",
          "blocks.6.d_write.write.proj.weight",
          "blocks.6.d_write.write.proj.bias",
          "blocks.6.d_write.allocation.weight",
          "blocks.6.d_write.allocation.bias",
          "blocks.6.d_read.residual.raw_eta",
          "blocks.6.a_res.raw_eta",
          "blocks.6.c_read.residual.raw_eta",
          "blocks.6.f_res.raw_eta",
          "blocks.7.raw_gamma",
          "blocks.7.d_write.retention.proj.weight",
          "blocks.7.d_write.retention.proj.bias",
          "blocks.7.d_write.write.proj.weight",
          "blocks.7.d_write.write.proj.bias",
          "blocks.7.d_write.allocation.weight",
          "blocks.7.d_write.allocation.bias",
          "blocks.7.d_read.residual.raw_eta",
          "blocks.7.a_res.raw_eta",
          "blocks.7.c_read.residual.raw_eta",
          "blocks.7.f_res.raw_eta"
        ],
        "label": "c_write.retention.proj.weight, c_write.retention.proj.bias, c_write.write.proj.weight, c_write.write.proj.bias, c_write.threshold.weight, c_write.threshold.bias, blocks.*.raw_gamma, blocks.*.a_res.raw_eta, blocks.*.c_read.residual.raw_eta, blocks.*.f_res.raw_eta, blocks.*.d_write.retention.proj.weight, blocks.*.d_write.retention.proj.bias, blocks.*.d_write.write.proj.weight, blocks.*.d_write.write.proj.bias, blocks.*.d_write.allocation.weight, blocks.*.d_write.allocation.bias, blocks.*.d_read.residual.raw_eta"
      }
    ],
    "basis": "契約のparameter_settings（最終倍率）",
    "base_lrs": [
      0.0004,
      0.00035,
      0.0003
    ],
    "schedule": {
      "lr": [
        0.0004,
        0.00035,
        0.0003
      ],
      "warmup_words": 1600000,
      "boundaries": [
        30000000,
        60000000
      ]
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.95
      ],
      "eps": 1e-08
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 16384,
    "loss": "final pass B NLL only, A gradients connected",
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "peak_lr": 0.0008,
      "min_lr": 8e-05,
      "decay": "continuous_cosine_no_restarts",
      "dropout_boundaries_words": [
        0,
        10000000,
        30000000,
        100000000
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ]
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.95
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      512,
      512,
      512,
      512,
      512,
      512,
      512,
      512,
      512,
      512
    ],
    "batch": 16384,
    "loss": null,
    "initialization": null,
    "runtime_optimizations": null
  },
  {
    "groups": [],
    "basis": "",
    "base_lrs": [
      0.0008,
      0.0007,
      0.0006
    ],
    "schedule": {
      "coordinate": "raw_words_exposed",
      "warmup_words": 1600000,
      "boundaries": [
        0,
        30000000,
        60000000,
        100000000
      ],
      "lr": [
        0.0008,
        0.0007,
        0.0006
      ],
      "ffn_dropout": [
        0.0,
        0.02,
        0.05
      ],
      "decay": "none"
    },
    "optimizer": {
      "name": "AdamW",
      "betas": [
        0.9,
        0.9872585449014338
      ],
      "eps": 1e-08,
      "matrix_weight_decay": 0.1,
      "vector_weight_decay": 0.0,
      "gradient_clip": 1.0,
      "fused": true
    },
    "sequence": [
      128,
      256,
      512,
      128,
      256,
      512,
      128,
      256,
      512,
      512
    ],
    "batch": 4096,
    "loss": "Final CE only; no auxiliary loss",
    "initialization": "Exact common tensors and RNG/cursor from highest-score272 zero-word full-state;72 fresh calibrated gains;empty Adam rebuilt with identical width-specific LR/WD rules",
    "runtime_optimizations": {
      "diagnostic_every_updates": 25,
      "host_gather_indices": "sorted identical indices from packed lengths/segments; gradient equality required",
      "evaluation": "FP32 exact chunked eager CUDA graph, validated prompt pooling"
    }
  }
]