{
  "scope": "原创教学数据与CPU标量算术。没有商业LLM调用、真实语义训练或GPU性能测量。",
  "vocabulary": [
    "<PAD>",
    "我",
    "爱",
    "学习",
    "你",
    "！"
  ],
  "embedding_table": [
    [
      0.0,
      0.0,
      0.0,
      0.0
    ],
    [
      0.2,
      -0.1,
      0.7,
      0.4
    ],
    [
      -0.3,
      0.8,
      0.1,
      0.5
    ],
    [
      0.6,
      0.2,
      -0.4,
      0.9
    ],
    [
      0.1,
      -0.5,
      0.9,
      0.3
    ],
    [
      -0.2,
      0.4,
      0.5,
      -0.6
    ]
  ],
  "input_ids": [
    [
      1,
      2,
      3
    ],
    [
      4,
      2,
      0
    ]
  ],
  "valid_positions": [
    [
      1,
      1,
      1
    ],
    [
      1,
      1,
      0
    ]
  ],
  "id_shape": [
    2,
    3
  ],
  "embedding_output": [
    [
      [
        0.2,
        -0.1,
        0.7,
        0.4
      ],
      [
        -0.3,
        0.8,
        0.1,
        0.5
      ],
      [
        0.6,
        0.2,
        -0.4,
        0.9
      ]
    ],
    [
      [
        0.1,
        -0.5,
        0.9,
        0.3
      ],
      [
        -0.3,
        0.8,
        0.1,
        0.5
      ],
      [
        0.0,
        0.0,
        0.0,
        0.0
      ]
    ]
  ],
  "embedding_shape": [
    2,
    3,
    4
  ],
  "shape_scale": {
    "B": 2,
    "T": 128,
    "d": 4096,
    "elements": 1048576,
    "bytes_per_element": 2,
    "output_bytes": 2097152,
    "output_MiB": 2.0,
    "input_int64_bytes": 2048
  },
  "matrix_product": [
    [
      4,
      5
    ],
    [
      10,
      11
    ]
  ],
  "transpose": [
    [
      1,
      4
    ],
    [
      2,
      5
    ],
    [
      3,
      6
    ]
  ],
  "reshape": [
    [
      1,
      2
    ],
    [
      3,
      4
    ],
    [
      5,
      6
    ]
  ],
  "softmax_ln123": [
    0.16666666666666666,
    0.3333333333333333,
    0.5
  ],
  "attention": {
    "scores": [
      0.7071067811865475,
      0.0
    ],
    "weights": [
      0.6697615493266569,
      0.3302384506733431
    ],
    "output": [
      6.697615493266569,
      6.604769013466862
    ]
  },
  "mask_without_cache": [
    [
      true,
      false,
      false
    ],
    [
      true,
      true,
      false
    ],
    [
      true,
      true,
      true
    ]
  ],
  "mask_3_past_2_new": [
    [
      true,
      true,
      true,
      true,
      false
    ],
    [
      true,
      true,
      true,
      true,
      true
    ]
  ],
  "rotation_90": [
    -3.0,
    2.0
  ],
  "layernorm_no_epsilon": [
    -1.0,
    1.0
  ],
  "rmsnorm_no_epsilon": [
    0.4472135954999579,
    1.3416407864998738
  ],
  "output_logits": [
    0,
    1,
    2,
    3,
    -1,
    -2
  ],
  "next_token_greedy": 3,
  "simple_network": {
    "preactivation": [
      3,
      -1
    ],
    "relu": [
      3,
      0
    ],
    "output": 6
  },
  "kv_toy": {
    "layers": 2,
    "positions": 3,
    "heads": 1,
    "head_width": 4,
    "bytes_per_element": 4,
    "elements": 48,
    "bytes": 192
  },
  "scalar_training": {
    "w": 1.0,
    "x": 2.0,
    "target": 4.0,
    "prediction": 2.0,
    "loss": 2.0,
    "gradient": -4.0,
    "learning_rate": 0.2,
    "new_w": 1.8,
    "new_prediction": 3.6,
    "new_loss": 0.07999999999999996
  },
  "finite_difference_gradient": -4.000000000226045,
  "dpo_example": {
    "score": 0.6931471805599453,
    "sigmoid": 0.6666666666666666,
    "loss": 0.4054651081081644
  },
  "kl_example": 0.14384103622589042,
  "rrf": {
    "A": 0.16783216783216784,
    "B": 0.17424242424242425,
    "C": 0.16025641025641024
  },
  "pass_independent_p08_k3": {
    "at_least_once": 0.992,
    "all_three": 0.5120000000000001
  },
  "dag": {
    "serial": 10,
    "parallel_ideal": 7,
    "parallel_overhead4": 11
  },
  "image_patches": {
    "height": 224,
    "width": 224,
    "patch_side": 16,
    "count": 196,
    "raw_patch_values": 768
  }
}