Rankine–Hugoniot Front Tokens / bench_report.json

Failed on benchmark

Raw ⬇ ZIP
  1{
  2  "bench_version": 1,
  3  "track": "burgers_periodic_split",
  4  "model": "cnn_shared",
  5  "metric_direction": "lower is better",
  6  "n_seeds": 8,
  7  "baseline": {
  8    "best_cfg": {
  9      "lr": 0.006,
 10      "epochs": 15
 11    },
 12    "sweep": [
 13      {
 14        "cfg": {
 15          "lr": 0.0015,
 16          "epochs": 15
 17        },
 18        "mean": 0.0002808531244227197
 19      },
 20      {
 21        "cfg": {
 22          "lr": 0.003,
 23          "epochs": 15
 24        },
 25        "mean": 0.0002247071788588073
 26      },
 27      {
 28        "cfg": {
 29          "lr": 0.006,
 30          "epochs": 15
 31        },
 32        "mean": 0.00017973236390389502
 33      }
 34    ],
 35    "full": {
 36      "mean": 0.0001909858247017837,
 37      "std": 4.899876337576311e-05,
 38      "per_seed": [
 39        0.00019845497445203364,
 40        0.00017688231309875846,
 41        0.00018345325952395797,
 42        0.00016013890854083002,
 43        0.00018341024406254292,
 44        0.00011085406731581315,
 45        0.000293802673695609,
 46        0.00022089015692472458
 47      ],
 48      "n": 8
 49    }
 50  },
 51  "idea": {
 52    "mean": 0.0022266772139118984,
 53    "std": 0.0008782478931862389,
 54    "per_seed": [
 55      0.002027727896347642,
 56      0.0031949642580002546,
 57      0.0016515825409442186,
 58      0.0014884979464113712,
 59      0.0014188691275194287,
 60      0.001367701217532158,
 61      0.00382382795214653,
 62      0.0028402467723935843
 63    ],
 64    "n": 8,
 65    "selected_cfg": {
 66      "lr": 0.006,
 67      "epochs": 15
 68    }
 69  },
 70  "comparison": {
 71    "delta_mean": 0.0020356913892101147,
 72    "idea_wins": 0,
 73    "n_pairs": 8,
 74    "per_seed_diffs": [
 75      0.0018292729218956083,
 76      0.003018081944901496,
 77      0.0014681292814202607,
 78      0.0013283590378705412,
 79      0.0012354588834568858,
 80      0.0012568471502163447,
 81      0.003530025278450921,
 82      0.0026193566154688597
 83    ],
 84    "p_value": 0.0081,
 85    "mde": 0.000751264160788154,
 86    "mde_rel_pct": 393.36121513794086,
 87    "verdict": "idea worse (significant)",
 88    "system_worked": false
 89  },
 90  "mechanism_signature": {
 91    "idea_sweep": [
 92      {
 93        "cfg": {
 94          "lr": 0.0015,
 95          "epochs": 15
 96        },
 97        "mean": 0.010121025261469185
 98      },
 99      {
100        "cfg": {
101          "lr": 0.003,
102          "epochs": 15
103        },
104        "mean": 0.0046673151082359254
105      },
106      {
107        "cfg": {
108          "lr": 0.006,
109          "epochs": 15
110        },
111        "mean": 0.0020906931604258716
112      }
113    ],
114    "custom_track": {
115      "name": "burgers_front_tokens_bench",
116      "file": "bench_front_tokens.py",
117      "domain": "pde"
118    },
119    "mechanism_signature": {
120      "predicted_mean_speed_displacement": -0.00025935691202875956,
121      "observed_mean_displacement": 0.0,
122      "relative_error": 259.35691202875955,
123      "confirmed": false
124    }
125  }
126}