data/audio_atlas.json

{
  "n_models": 6,
  "n_pairs": 15,
  "models": [
    {
      "name": "hubert-base",
      "family": "hubert",
      "objective": "self-supervised",
      "dim": 768
    },
    {
      "name": "hubert-large",
      "family": "hubert",
      "objective": "self-supervised",
      "dim": 1024
    },
    {
      "name": "wav2vec2-base",
      "family": "wav2vec2",
      "objective": "self-supervised",
      "dim": 768
    },
    {
      "name": "wav2vec2-large",
      "family": "wav2vec2",
      "objective": "self-supervised",
      "dim": 1024
    },
    {
      "name": "whisper-base",
      "family": "whisper",
      "objective": "supervised",
      "dim": 512
    },
    {
      "name": "whisper-small",
      "family": "whisper",
      "objective": "supervised",
      "dim": 768
    }
  ],
  "regime_counts": {
    "Local-Only": 14,
    "Convergent": 1
  },
  "pairs": [
    {
      "pair": "hubert-base \u00d7 hubert-large",
      "model_a": "hubert-base",
      "model_b": "hubert-large",
      "family_a": "hubert",
      "family_b": "hubert",
      "objective_a": "self-supervised",
      "objective_b": "self-supervised",
      "s_local": 0.465,
      "s_local_raw": 0.4677,
      "null_mean": 0.005,
      "s_global": 0.778,
      "lambda": 0.5978,
      "tau": 0.0,
      "tau_a_to_b": 0.9469,
      "tau_b_to_a": 0.0,
      "tai": 0.2126,
      "bitrate_a": 5.92,
      "bitrate_b": 7.1,
      "delta": 1.18,
      "regime": "Local-Only"
    },
    {
      "pair": "hubert-base \u00d7 wav2vec2-base",
      "model_a": "hubert-base",
      "model_b": "wav2vec2-base",
      "family_a": "hubert",
      "family_b": "wav2vec2",
      "objective_a": "self-supervised",
      "objective_b": "self-supervised",
      "s_local": 0.3995,
      "s_local_raw": 0.4025,
      "null_mean": 0.005,
      "s_global": 0.6982,
      "lambda": 0.5722,
      "tau": 0.0,
      "tau_a_to_b": 0.0,
      "tau_b_to_a": 0.9042,
      "tai": 0.2619,
      "bitrate_a": 5.92,
      "bitrate_b": 5.37,
      "delta": 0.55,
      "regime": "Local-Only"
    },
    {
      "pair": "hubert-base \u00d7 wav2vec2-large",
      "model_a": "hubert-base",
      "model_b": "wav2vec2-large",
      "family_a": "hubert",
      "family_b": "wav2vec2",
      "objective_a": "self-supervised",
      "objective_b": "self-supervised",
      "s_local": 0.29,
      "s_local_raw": 0.2936,
      "null_mean": 0.005,
      "s_global": 0.7394,
      "lambda": 0.3922,
      "tau": 0.0,
      "tau_a_to_b": 0.0,
      "tau_b_to_a": 0.2579,
      "tai": 0.49,
      "bitrate_a": 5.92,
      "bitrate_b": 3.12,
      "delta": 2.8,
      "regime": "Local-Only"
    },
    {
      "pair": "hubert-base \u00d7 whisper-base",
      "model_a": "hubert-base",
      "model_b": "whisper-base",
      "family_a": "hubert",
      "family_b": "whisper",
      "objective_a": "self-supervised",
      "objective_b": "supervised",
      "s_local": 0.1758,
      "s_local_raw": 0.1799,
      "null_mean": 0.005,
      "s_global": 0.6128,
      "lambda": 0.2868,
      "tau": 0.0,
      "tau_a_to_b": 0.0,
      "tau_b_to_a": 0.6392,
      "tai": 0.3776,
      "bitrate_a": 5.92,
      "bitrate_b": 2.92,
      "delta": 2.99,
      "regime": "Local-Only"
    },
    {
      "pair": "hubert-base \u00d7 whisper-small",
      "model_a": "hubert-base",
      "model_b": "whisper-small",
      "family_a": "hubert",
      "family_b": "whisper",
      "objective_a": "self-supervised",
      "objective_b": "supervised",
      "s_local": 0.1536,
      "s_local_raw": 0.1578,
      "null_mean": 0.005,
      "s_global": 0.6418,
      "lambda": 0.2393,
      "tau": 0.0,
      "tau_a_to_b": 0.0,
      "tau_b_to_a": 0.5425,
      "tai": 0.2367,
      "bitrate_a": 5.92,
      "bitrate_b": 3.31,
      "delta": 2.61,
      "regime": "Local-Only"
    },
    {
      "pair": "hubert-large \u00d7 wav2vec2-base",
      "model_a": "hubert-large",
      "model_b": "wav2vec2-base",
      "family_a": "hubert",
      "family_b": "wav2vec2",
      "objective_a": "self-supervised",
      "objective_b": "self-supervised",
      "s_local": 0.3285,
      "s_local_raw": 0.3318,
      "null_mean": 0.005,
      "s_global": 0.5825,
      "lambda": 0.5639,
      "tau": 0.0,
      "tau_a_to_b": 0.0,
      "tau_b_to_a": 1.0,
      "tai": 0.757,
      "bitrate_a": 7.1,
      "bitrate_b": 5.37,
      "delta": 1.73,
      "regime": "Local-Only"
    },
    {
      "pair": "hubert-large \u00d7 wav2vec2-large",
      "model_a": "hubert-large",
      "model_b": "wav2vec2-large",
      "family_a": "hubert",
      "family_b": "wav2vec2",
      "objective_a": "self-supervised",
      "objective_b": "self-supervised",
      "s_local": 0.2244,
      "s_local_raw": 0.2283,
      "null_mean": 0.005,
      "s_global": 0.5251,
      "lambda": 0.4274,
      "tau": 0.0,
      "tau_a_to_b": 0.0,
      "tau_b_to_a": 0.4506,
      "tai": 0.8781,
      "bitrate_a": 7.1,
      "bitrate_b": 3.12,
      "delta": 3.98,
      "regime": "Local-Only"
    },
    {
      "pair": "hubert-large \u00d7 whisper-base",
      "model_a": "hubert-large",
      "model_b": "whisper-base",
      "family_a": "hubert",
      "family_b": "whisper",
      "objective_a": "self-supervised",
      "objective_b": "supervised",
      "s_local": 0.1444,
      "s_local_raw": 0.1487,
      "null_mean": 0.005,
      "s_global": 0.4495,
      "lambda": 0.3213,
      "tau": 0.0,
      "tau_a_to_b": 0.0,
      "tau_b_to_a": 0.4336,
      "tai": 0.5493,
      "bitrate_a": 7.1,
      "bitrate_b": 2.92,
      "delta": 4.17,
      "regime": "Local-Only"
    },
    {
      "pair": "hubert-large \u00d7 whisper-small",
      "model_a": "hubert-large",
      "model_b": "whisper-small",
      "family_a": "hubert",
      "family_b": "whisper",
      "objective_a": "self-supervised",
      "objective_b": "supervised",
      "s_local": 0.1327,
      "s_local_raw": 0.137,
      "null_mean": 0.005,
      "s_global": 0.4649,
      "lambda": 0.2855,
      "tau": 0.0,
      "tau_a_to_b": 0.0,
      "tau_b_to_a": 0.3516,
      "tai": 0.4067,
      "bitrate_a": 7.1,
      "bitrate_b": 3.31,
      "delta": 3.79,
      "regime": "Local-Only"
    },
    {
      "pair": "wav2vec2-base \u00d7 wav2vec2-large",
      "model_a": "wav2vec2-base",
      "model_b": "wav2vec2-large",
      "family_a": "wav2vec2",
      "family_b": "wav2vec2",
      "objective_a": "self-supervised",
      "objective_b": "self-supervised",
      "s_local": 0.3109,
      "s_local_raw": 0.3144,
      "null_mean": 0.005,
      "s_global": 0.6808,
      "lambda": 0.4567,
      "tau": 0.0,
      "tau_a_to_b": 0.0,
      "tau_b_to_a": 0.2416,
      "tai": 0.2506,
      "bitrate_a": 5.37,
      "bitrate_b": 3.12,
      "delta": 2.25,
      "regime": "Local-Only"
    },
    {
      "pair": "wav2vec2-base \u00d7 whisper-base",
      "model_a": "wav2vec2-base",
      "model_b": "whisper-base",
      "family_a": "wav2vec2",
      "family_b": "whisper",
      "objective_a": "self-supervised",
      "objective_b": "supervised",
      "s_local": 0.1398,
      "s_local_raw": 0.1442,
      "null_mean": 0.005,
      "s_global": 0.4896,
      "lambda": 0.2856,
      "tau": 0.0,
      "tau_a_to_b": 0.0,
      "tau_b_to_a": 0.6632,
      "tai": 0.3051,
      "bitrate_a": 5.37,
      "bitrate_b": 2.92,
      "delta": 2.44,
      "regime": "Local-Only"
    },
    {
      "pair": "wav2vec2-base \u00d7 whisper-small",
      "model_a": "wav2vec2-base",
      "model_b": "whisper-small",
      "family_a": "wav2vec2",
      "family_b": "whisper",
      "objective_a": "self-supervised",
      "objective_b": "supervised",
      "s_local": 0.1272,
      "s_local_raw": 0.1316,
      "null_mean": 0.005,
      "s_global": 0.5278,
      "lambda": 0.2409,
      "tau": 0.0,
      "tau_a_to_b": 0.0,
      "tau_b_to_a": 0.5144,
      "tai": 0.0109,
      "bitrate_a": 5.37,
      "bitrate_b": 3.31,
      "delta": 2.06,
      "regime": "Local-Only"
    },
    {
      "pair": "wav2vec2-large \u00d7 whisper-base",
      "model_a": "wav2vec2-large",
      "model_b": "whisper-base",
      "family_a": "wav2vec2",
      "family_b": "whisper",
      "objective_a": "self-supervised",
      "objective_b": "supervised",
      "s_local": 0.1415,
      "s_local_raw": 0.1458,
      "null_mean": 0.005,
      "s_global": 0.6056,
      "lambda": 0.2337,
      "tau": 0.0,
      "tau_a_to_b": 0.4827,
      "tau_b_to_a": 0.0,
      "tai": 0.1978,
      "bitrate_a": 3.12,
      "bitrate_b": 2.92,
      "delta": 0.19,
      "regime": "Local-Only"
    },
    {
      "pair": "wav2vec2-large \u00d7 whisper-small",
      "model_a": "wav2vec2-large",
      "model_b": "whisper-small",
      "family_a": "wav2vec2",
      "family_b": "whisper",
      "objective_a": "self-supervised",
      "objective_b": "supervised",
      "s_local": 0.1357,
      "s_local_raw": 0.14,
      "null_mean": 0.005,
      "s_global": 0.6602,
      "lambda": 0.2055,
      "tau": 0.0,
      "tau_a_to_b": 0.4903,
      "tau_b_to_a": 0.0,
      "tai": 0.3098,
      "bitrate_a": 3.12,
      "bitrate_b": 3.31,
      "delta": 0.19,
      "regime": "Local-Only"
    },
    {
      "pair": "whisper-base \u00d7 whisper-small",
      "model_a": "whisper-base",
      "model_b": "whisper-small",
      "family_a": "whisper",
      "family_b": "whisper",
      "objective_a": "supervised",
      "objective_b": "supervised",
      "s_local": 0.2699,
      "s_local_raw": 0.2735,
      "null_mean": 0.005,
      "s_global": 0.9104,
      "lambda": 0.2964,
      "tau": 0.8785,
      "tau_a_to_b": 0.9275,
      "tau_b_to_a": 0.8785,
      "tai": 0.0504,
      "bitrate_a": 2.92,
      "bitrate_b": 3.31,
      "delta": 0.38,
      "regime": "Convergent"
    }
  ]
}