# Стенд-проверка (stage-2) · промт оператора:

(универсальный)

Ответ агента:

{
  "worked": true,
  "confidence": 8,
  "verdict": "The registered custom track resolution_hilbert_regression was rerun through bench.get_dataset and bench.make_report with eight paired seeds and a parity-matched three-value learning-rate sweep. The constrained Hilbert dictionary achieved lower test MSE than the tuned unconstrained baseline in all eight pairs: 0.06969587 versus 0.07284840, delta=-0.00315253, permutation p=0.0081. The trained-model mechanism signature confirmed the projected hidden-weight bound and output l1 constraint.",
  "metrics": {
    "baseline": "Best lr=0.003; mean test MSE 0.0728484001; sweep means [0.0833553793, 0.0728484001, 0.0748559255].",
    "idea": "Best lr=0.003; mean test MSE 0.0696958704; per-seed results [0.0712435767, 0.0731648430, 0.0639909878, 0.0699851513, 0.0695066899, 0.0629635379, 0.0717324987, 0.0749796778]."
  },
  "bench_report": {
    "bench_version": 1,
    "track": "resolution_hilbert_regression",
    "model": "local_dictionary",
    "metric_direction": "lower is better",
    "n_seeds": 8,
    "baseline": {
      "best_cfg": {
        "lr": 0.003,
        "weight_decay": 0.0
      },
      "sweep": [
        {
          "cfg": {
            "lr": 0.0015,
            "weight_decay": 0.0
          },
          "mean": 0.08335537929087877
        },
        {
          "cfg": {
            "lr": 0.003,
            "weight_decay": 0.0
          },
          "mean": 0.07284840010106564
        },
        {
          "cfg": {
            "lr": 0.006,
            "weight_decay": 0.0
          },
          "mean": 0.07485592551529408
        }
      ],
      "full": {
        "mean": 0.07284840010106564,
        "std": 0.003659818625431143,
        "per_seed": [
          0.07480311393737793,
          0.07678214460611343,
          0.0665731206536293,
          0.07224995642900467,
          0.07274671643972397,
          0.06788084656000137,
          0.07427780330181122,
          0.07747349888086319
        ],
        "n": 8
      }
    },
    "idea": {
      "mean": 0.06969587039202452,
      "std": 0.003944697395537844,
      "per_seed": [
        0.07124357670545578,
        0.07316484302282333,
        0.0639909878373146,
        0.06998515129089355,
        0.0695066899061203,
        0.06296353787183762,
        0.07173249870538712,
        0.07497967779636383
      ],
      "n": 8
    },
    "comparison": {
      "delta_mean": -0.0031525297090411186,
      "idea_wins": 8,
      "n_pairs": 8,
      "per_seed_diffs": [
        -0.0035595372319221497,
        -0.0036173015832901,
        -0.0025821328163146973,
        -0.0022648051381111145,
        -0.003240026533603668,
        -0.004917308688163757,
        -0.002545304596424103,
        -0.002493821084499359
      ],
      "p_value": 0.0081,
      "mde": 0.0007357471367279225,
      "mde_rel_pct": 1.0099702062189282,
      "verdict": "idea better (significant)",
      "system_worked": true
    },
    "mechanism_signature": {
      "custom_track": {
        "name": "resolution_hilbert_regression",
        "file": "resolution_track.py",
        "domain": "function_measurements"
      },
      "confirmed": true,
      "observed_mean_max_weight_norm": 1.0000000447,
      "observed_mean_output_l1": 2.4546456635
    }
  },
  "how_to_run": "/home/maxwelhelp/main/bin/python3 bench_experiment.py",
  "files": [
    "bench_experiment.py",
    "resolution_track.py",
    "bench_report.json"
  ],
  "limitations": "The registered custom task is a fast synthetic function-measurement regression benchmark rather than Burgers, Darcy, or a real measured operator dataset. Distinct-resolution transfer, FLOP/wall-clock matching, larger widths, and alternative basis normalizations were not tested.",
  "system_verdict": "worked",
  "practical_verdict": "helps",
  "mechanism_ok": 1,
  "system_judged": true
}