{
  "bench_version": "local_fallback_no_harness",
  "track": "dynamics",
  "model": "shared DelayGRU",
  "metric_direction": "lower is better",
  "n_seeds": 8,
  "baseline": {
    "best_cfg": {
      "lr": 0.006,
      "wd": 0.0
    },
    "sweep": [
      {
        "cfg": {
          "lr": 0.001,
          "wd": 0.0
        },
        "mean": 0.6801233738660812
      },
      {
        "cfg": {
          "lr": 0.001,
          "wd": 0.0001
        },
        "mean": 0.6801258623600006
      },
      {
        "cfg": {
          "lr": 0.003,
          "wd": 0.0
        },
        "mean": 0.35754771530628204
      },
      {
        "cfg": {
          "lr": 0.003,
          "wd": 0.0001
        },
        "mean": 0.3575517013669014
      },
      {
        "cfg": {
          "lr": 0.006,
          "wd": 0.0
        },
        "mean": 0.16513000801205635
      },
      {
        "cfg": {
          "lr": 0.006,
          "wd": 0.0001
        },
        "mean": 0.16513226553797722
      }
    ],
    "full": {
      "mean": 0.17238342948257923,
      "std": 0.023754673289860206,
      "per_seed": [
        0.15474747121334076,
        0.1601465791463852,
        0.16793453693389893,
        0.17769144475460052,
        0.16201430559158325,
        0.2271566540002823,
        0.15431520342826843,
        0.17506124079227448
      ],
      "n": 8
    }
  },
  "idea": {
    "mean": 0.17108880542218685,
    "std": 0.022775696321881225,
    "per_seed": [
      0.15279607474803925,
      0.1616012454032898,
      0.16678939759731293,
      0.1741822063922882,
      0.15894076228141785,
      0.2237611562013626,
      0.15557505190372467,
      0.1750645488500595
    ],
    "n": 8,
    "best_cfg": {
      "lr": 0.006,
      "wd": 0.0
    },
    "grid": [
      {
        "cfg": {
          "lr": 0.001,
          "wd": 0.0
        },
        "mean": 0.6519463434815407
      },
      {
        "cfg": {
          "lr": 0.003,
          "wd": 0.0
        },
        "mean": 0.3875108100473881
      },
      {
        "cfg": {
          "lr": 0.006,
          "wd": 0.0
        },
        "mean": 0.17108880542218685
      }
    ]
  },
  "comparison": {
    "delta_mean": -0.0012946240603923798,
    "idea_wins": 5,
    "n_pairs": 8,
    "per_seed_diffs": [
      -0.0019513964653015137,
      0.001454666256904602,
      -0.0011451393365859985,
      -0.003509238362312317,
      -0.0030735433101654053,
      -0.0033954977989196777,
      0.0012598484754562378,
      3.3080577850341797e-06
    ],
    "p_value": 0.1247937603119844,
    "verdict": "idea better (not significant)",
    "system_worked": false
  },
  "mechanism_signature": {
    "prediction": "reconstruction/prediction error grows approximately linearly with observation noise (inverse observability intuition)",
    "observed_loglog_slopes": {
      "baseline": 0.0070740981333970365,
      "idea": 0.007060657411190435
    },
    "per_noise": [
      {
        "sigma": 0.01,
        "baseline_rmse": 0.3934699594974518,
        "idea_rmse": 0.3909880518913269
      },
      {
        "sigma": 0.03,
        "baseline_rmse": 0.394115686416626,
        "idea_rmse": 0.3916374146938324
      },
      {
        "sigma": 0.06,
        "baseline_rmse": 0.3962407410144806,
        "idea_rmse": 0.3937460780143738
      },
      {
        "sigma": 0.1,
        "baseline_rmse": 0.40119439363479614,
        "idea_rmse": 0.398638516664505
      },
      {
        "sigma": 0.01,
        "baseline_rmse": 0.4001429080963135,
        "idea_rmse": 0.4019409716129303
      },
      {
        "sigma": 0.03,
        "baseline_rmse": 0.40048766136169434,
        "idea_rmse": 0.40225452184677124
      },
      {
        "sigma": 0.06,
        "baseline_rmse": 0.4020640552043915,
        "idea_rmse": 0.40378326177597046
      },
      {
        "sigma": 0.1,
        "baseline_rmse": 0.40611129999160767,
        "idea_rmse": 0.4077668786048889
      },
      {
        "sigma": 0.01,
        "baseline_rmse": 0.4102720320224762,
        "idea_rmse": 0.40883108973503113
      },
      {
        "sigma": 0.03,
        "baseline_rmse": 0.41178008913993835,
        "idea_rmse": 0.41025447845458984
      },
      {
        "sigma": 0.06,
        "baseline_rmse": 0.4154220521450043,
        "idea_rmse": 0.41376587748527527
      },
      {
        "sigma": 0.1,
        "baseline_rmse": 0.42276376485824585,
        "idea_rmse": 0.4209310710430145
      },
      {
        "sigma": 0.01,
        "baseline_rmse": 0.421396940946579,
        "idea_rmse": 0.4172685146331787
      },
      {
        "sigma": 0.03,
        "baseline_rmse": 0.4214303195476532,
        "idea_rmse": 0.4174094498157501
      },
      {
        "sigma": 0.06,
        "baseline_rmse": 0.4222506880760193,
        "idea_rmse": 0.41838502883911133
      },
      {
        "sigma": 0.1,
        "baseline_rmse": 0.4247669279575348,
        "idea_rmse": 0.42109447717666626
      }
    ],
    "confirmed": false
  },
  "protocol_note": "Official /home/maxwelhelp/all/math2nn/bench and README were absent; this is a local fallback and is not official harness evidence."
}