data/e4_gamma/EleutherAI--pythia-31m_mongo.json

{
  "model": "EleutherAI/pythia-31m",
  "theta": 10000,
  "gamma_pred": 0.99999561666838,
  "C_theory": 9.2103,
  "distances": [
    10,
    20,
    30,
    50,
    100,
    200,
    500,
    1000,
    2000
  ],
  "distances_fit": [
    30,
    50,
    100,
    200,
    500,
    1000,
    2000
  ],
  "d_min_fit": 30,
  "n_prompts_per_distance": 150,
  "seeds": [
    42,
    123,
    7
  ],
  "attn_by_distance": {
    "10": {
      "mean_across_seeds": 0.015075270254164932,
      "std_across_seeds": 0.0016076363765655506,
      "per_seed_means": [
        0.015652700106923777,
        0.01288217271057268,
        0.016690937944998343
      ]
    },
    "20": {
      "mean_across_seeds": 0.007979718983923603,
      "std_across_seeds": 0.0008011809172800115,
      "per_seed_means": [
        0.007325330031647657,
        0.007505871877074241,
        0.009107955043048909
      ]
    },
    "30": {
      "mean_across_seeds": 0.004943025353131816,
      "std_across_seeds": 0.0008214856215946895,
      "per_seed_means": [
        0.004706252690715095,
        0.004076418298839902,
        0.0060464050698404515
      ]
    },
    "50": {
      "mean_across_seeds": 0.003909375242526746,
      "std_across_seeds": 0.0004103009057058668,
      "per_seed_means": [
        0.004045853732774655,
        0.0033527197137785455,
        0.004329552281027039
      ]
    },
    "100": {
      "mean_across_seeds": 0.001335802334417369,
      "std_across_seeds": 0.00014613574339847555,
      "per_seed_means": [
        0.0012400429498666198,
        0.0012250753167124156,
        0.0015422887366730719
      ]
    },
    "200": {
      "mean_across_seeds": 0.0006274800656319712,
      "std_across_seeds": 8.133098864096775e-05,
      "per_seed_means": [
        0.000617688476195326,
        0.0005331277462513148,
        0.0007316239744492729
      ]
    },
    "500": {
      "mean_across_seeds": 0.0003229697247034993,
      "std_across_seeds": 0.00010411438760957841,
      "per_seed_means": [
        0.0002789432165203228,
        0.0002233032413945087,
        0.00046666271619566637
      ]
    },
    "1000": {
      "mean_across_seeds": 0.00010229767169578944,
      "std_across_seeds": 6.545994570499704e-05,
      "per_seed_means": [
        0.00014693857560094632,
        9.742619926242924e-06,
        0.00015021181956017908
      ]
    },
    "2000": {
      "mean_across_seeds": 2.207666448479115e-05,
      "std_across_seeds": 9.890072727231223e-06,
      "per_seed_means": [
        3.0325915145184013e-05,
        8.170285452860545e-06,
        2.7733792856328893e-05
      ]
    }
  },
  "fit_power_law": {
    "gamma": 1.2350013988825523,
    "log_A": -0.8481173688844952,
    "R2": 0.973742,
    "n_points": 7
  },
  "fit_exponential": {
    "lambda": 0.0025245756131658517,
    "log_A": -6.1424841272489665,
    "R2": 0.876899,
    "n_points": 7
  },
  "delta_R2_power_minus_exp": 0.0968,
  "gamma_CI_95": {
    "lo": 1.0142831942090689,
    "hi": 1.3931415910718121
  },
  "decision": "ANOMALY: long-context training effect"
}