Measurement-Space Neural Operator with Mesh Transfer / bench_report.json

Mechanism confirmed, baseline not beaten

Raw ⬇ ZIP
  1{
  2  "bench_version": 1,
  3  "track": "poisson_measurements",
  4  "model": "custom_measurement_mlp",
  5  "metric_direction": "lower is better",
  6  "n_seeds": 8,
  7  "baseline": {
  8    "best_cfg": {
  9      "lr": 0.01
 10    },
 11    "sweep": [
 12      {
 13        "cfg": {
 14          "lr": 0.001
 15        },
 16        "mean": 0.00023615473219251726
 17      },
 18      {
 19        "cfg": {
 20          "lr": 0.003
 21        },
 22        "mean": 8.562382845411776e-05
 23      },
 24      {
 25        "cfg": {
 26          "lr": 0.01
 27        },
 28        "mean": 7.791408143020817e-05
 29      }
 30    ],
 31    "full": {
 32      "mean": 7.791408143020817e-05,
 33      "std": 1.909924833468627e-05,
 34      "per_seed": [
 35        0.00011997614637948573,
 36        5.3765939810546115e-05,
 37        8.39241110952571e-05,
 38        6.556373409694061e-05,
 39        8.531597995897755e-05,
 40        8.173154492396861e-05,
 41        7.142309914343059e-05,
 42        6.161209603305906e-05
 43      ],
 44      "n": 8
 45    }
 46  },
 47  "idea": {
 48    "mean": 0.0148575731087476,
 49    "std": 0.0016423097465859962,
 50    "per_seed": [
 51      0.017279266317685445,
 52      0.014548975912233194,
 53      0.015849430797000727,
 54      0.017124839189151923,
 55      0.013294871213535469,
 56      0.014724781985084217,
 57      0.012648728614052137,
 58      0.013389690841237704
 59    ],
 60    "n": 8
 61  },
 62  "comparison": {
 63    "delta_mean": 0.014779659027317392,
 64    "idea_wins": 0,
 65    "n_pairs": 8,
 66    "per_seed_diffs": [
 67      0.01715929017130596,
 68      0.014495209972422648,
 69      0.01576550668590547,
 70      0.017059275455054982,
 71      0.013209555233576491,
 72      0.014643050440160248,
 73      0.012577305514908706,
 74      0.013328078745204645
 75    ],
 76    "p_value": 0.0081,
 77    "mde": 0.0014603972953644863,
 78    "mde_rel_pct": 1874.3688798701203,
 79    "verdict": "idea worse (significant)",
 80    "system_worked": false
 81  },
 82  "mechanism_signature": {
 83    "prediction": "permutation invariance of sensor aggregation",
 84    "predicted_max_abs": 0.0,
 85    "observed_max_abs": 2.2351741790771484e-08,
 86    "confirmed": true
 87  },
 88  "custom_track": {
 89    "name": "poisson_measurements",
 90    "file": "custom_track.py",
 91    "domain": "pde"
 92  },
 93  "idea_sweep": [
 94    {
 95      "cfg": {
 96        "lr": 0.001
 97      },
 98      "mean": 0.01485852306553473
 99    },
100    {
101      "cfg": {
102        "lr": 0.003
103      },
104      "mean": 0.0148575731087476
105    },
106    {
107      "cfg": {
108        "lr": 0.01
109      },
110      "mean": 0.014859355776570736
111    }
112  ]
113}