{
  "bench_version": 1,
  "track": "dynamics",
  "model": "residual_rnn_shared",
  "metric_direction": "lower is better",
  "n_seeds": 8,
  "baseline": {
    "best_cfg": {
      "lr": 0.006,
      "dt": 0.25,
      "m": 0
    },
    "sweep": [
      {
        "cfg": {
          "lr": 0.0015,
          "dt": 0.25,
          "m": 0
        },
        "mean": 0.0012726996355922893
      },
      {
        "cfg": {
          "lr": 0.003,
          "dt": 0.25,
          "m": 0
        },
        "mean": 0.0005508951580850407
      },
      {
        "cfg": {
          "lr": 0.006,
          "dt": 0.25,
          "m": 0
        },
        "mean": 0.0002624734515848104
      }
    ],
    "full": {
      "mean": 0.00025840229682216886,
      "std": 9.502614806341116e-05,
      "per_seed": [
        0.00021290902805048972,
        0.00015823994181118906,
        0.00023945324937812984,
        0.00043929158709943295,
        0.00016651710029691458,
        0.0003356700181029737,
        0.000337076693540439,
        0.00017806075629778206
      ],
      "n": 8
    }
  },
  "idea": {
    "mean": 0.00041295221308246255,
    "std": 0.00010711652432851176,
    "per_seed": [
      0.000396863033529371,
      0.00026599428383633494,
      0.0004898399929516017,
      0.0006024055765010417,
      0.00039211593684740365,
      0.00029558956157416105,
      0.0005119965644553304,
      0.00034881275496445596
    ],
    "n": 8
  },
  "comparison": {
    "delta_mean": 0.00015454991626029368,
    "idea_wins": 1,
    "n_pairs": 8,
    "per_seed_diffs": [
      0.0001839540054788813,
      0.00010775434202514589,
      0.0002503867435734719,
      0.0001631139894016087,
      0.00022559883655048907,
      -4.008045652881265e-05,
      0.00017491987091489136,
      0.0001707519986666739
    ],
    "p_value": 0.0163,
    "mde": 7.473365296214683e-05,
    "mde_rel_pct": 28.92143525085543,
    "verdict": "idea worse (significant)",
    "system_worked": false
  },
  "idea_sweep": [
    {
      "cfg": {
        "lr": 0.006,
        "dt": 0.25,
        "m": 0
      },
      "result": {
        "mean": 0.00041295221308246255,
        "std": 0.00010711652432851176,
        "per_seed": [
          0.000396863033529371,
          0.00026599428383633494,
          0.0004898399929516017,
          0.0006024055765010417,
          0.00039211593684740365,
          0.00029558956157416105,
          0.0005119965644553304,
          0.00034881275496445596
        ],
        "n": 8
      }
    },
    {
      "cfg": {
        "lr": 0.0015,
        "dt": 0.25,
        "m": 1
      },
      "result": {
        "mean": 0.011022639810107648,
        "std": 0.0020031189115314024,
        "per_seed": [
          0.011321489699184895,
          0.008100657723844051,
          0.01179476361721754,
          0.01137582492083311,
          0.011110322549939156,
          0.010274713858962059,
          0.015252543613314629,
          0.008950802497565746
        ],
        "n": 8
      }
    },
    {
      "cfg": {
        "lr": 0.006,
        "dt": 0.25,
        "m": 1
      },
      "result": {
        "mean": 0.0010292112237948459,
        "std": 0.00026172876255944693,
        "per_seed": [
          0.0011666412465274334,
          0.00048204700578935444,
          0.001413801102899015,
          0.0010645093861967325,
          0.0009550628019496799,
          0.0008412864408455789,
          0.0012185709783807397,
          0.0010917708277702332
        ],
        "n": 8
      }
    }
  ],
  "custom_track": null,
  "selection_note": "m=1 idea; baseline sweep and idea sweep use identical lr union and epochs",
  "mechanism_signature": {
    "quantity": "trained radial derivative vs -a*r^(m+1)",
    "predicted_mean": -4.693806644385415,
    "observed_mean": -4.693806647272514,
    "mean_abs_error": 5.941758198397501e-07,
    "relative_abs_error": 1.4333428362199376e-07,
    "correlation": 0.9999999999999342,
    "n_samples": 22400,
    "hidden_norm_mean": 2.9794825143207397,
    "confirmed": true
  }
}