evidence/evaluation.json

{
  "profile": "opt4_wide_seed29",
  "task": "opt4",
  "scored_classes": [
    "O0",
    "O1",
    "O2",
    "O3"
  ],
  "sequence_accuracy": 0.842255,
  "balanced_accuracy": 0.830951,
  "binary_accuracy": 0.93883,
  "num_sequences": 116321,
  "num_binaries": 376,
  "recorded_test_accuracy": 0.8422554826729481,
  "recomputed_matches_recorded": true,
  "checkpoint": "results/hopper_release/opt4_wide_seed29/opt4",
  "result_json": {
    "accuracy": 0.8422554826729481,
    "balanced_accuracy": 0.8309510174499452,
    "confusion": [
      [
        37645,
        14,
        34,
        19
      ],
      [
        10,
        21157,
        1149,
        636
      ],
      [
        1,
        2449,
        17545,
        6190
      ],
      [
        1,
        1860,
        5986,
        21625
      ]
    ],
    "n": 116321
  },
  "evaluated_at": "2026-09-06T03:48:35"
}