{
  "bench_version": 1,
  "track": "dynamics",
  "model": "rnn_small",
  "metric_direction": "lower is better",
  "n_seeds": 8,
  "baseline": {
    "best_cfg": {
      "lr": 0.006
    },
    "sweep": [
      {
        "cfg": {
          "lr": 0.0015
        },
        "mean": 0.0019174260669387877
      },
      {
        "cfg": {
          "lr": 0.003
        },
        "mean": 0.0009721333481138572
      },
      {
        "cfg": {
          "lr": 0.006
        },
        "mean": 0.0007438920001732185
      }
    ],
    "full": {
      "mean": 0.0007792052056174725,
      "std": 0.00021144133531208117,
      "per_seed": [
        0.0007514121243730187,
        0.0005835454212501645,
        0.0005766521790064871,
        0.0010639582760632038,
        0.0008014270570129156,
        0.0006233751191757619,
        0.0006578490138053894,
        0.001175422454252839
      ],
      "n": 8
    }
  },
  "idea": {
    "mean": 0.024256352335214615,
    "std": 0.0038067872943047523,
    "per_seed": [
      0.021198168396949768,
      0.02363172546029091,
      0.030124139040708542,
      0.02928881347179413,
      0.02132338099181652,
      0.018241815268993378,
      0.024838322773575783,
      0.02540445327758789
    ],
    "n": 8
  },
  "comparison": {
    "delta_mean": 0.023477147129597142,
    "idea_wins": 0,
    "n_pairs": 8,
    "per_seed_diffs": [
      0.02044675627257675,
      0.023048180039040744,
      0.029547486861702055,
      0.028224855195730925,
      0.020521953934803605,
      0.017618440149817616,
      0.024180473759770393,
      0.02422903082333505
    ],
    "p_value": 0.0081,
    "mde": 0.0033576493476355656,
    "mde_rel_pct": 430.9069451063066,
    "verdict": "idea worse (significant)",
    "system_worked": false
  },
  "mechanism_signature": {
    "idea_config": {
      "lr": 0.003
    },
    "idea_sweep": [
      {
        "cfg": {
          "lr": 0.0015
        },
        "result": {
          "mean": 0.1092000212520361,
          "std": 0.032826994153617124,
          "per_seed": [
            0.12505441904067993,
            0.13431435823440552,
            0.10349666327238083,
            0.1606961041688919,
            0.060887403786182404,
            0.12317634373903275,
            0.05900014936923981,
            0.10697472840547562
          ],
          "n": 8
        }
      },
      {
        "cfg": {
          "lr": 0.003
        },
        "result": {
          "mean": 0.024256352335214615,
          "std": 0.0038067872943047523,
          "per_seed": [
            0.021198168396949768,
            0.02363172546029091,
            0.030124139040708542,
            0.02928881347179413,
            0.02132338099181652,
            0.018241815268993378,
            0.024838322773575783,
            0.02540445327758789
          ],
          "n": 8
        }
      },
      {
        "cfg": {
          "lr": 0.006
        },
        "result": {
          "mean": 0.028757888358086348,
          "std": 0.005178806657688538,
          "per_seed": [
            0.021722447127103806,
            0.029085084795951843,
            0.03460188955068588,
            0.038650788366794586,
            0.024053702130913734,
            0.02563602291047573,
            0.02774704061448574,
            0.028566131368279457
          ],
          "n": 8
        }
      }
    ],
    "mechanism_signature": {
      "prediction": "direct topology interpolation can transiently violate calibration accuracy; exact receding-horizon filtering accepts no violating step",
      "trained_baseline_unfiltered_violations_of_9": 7,
      "trained_idea_accepted_step_violations": 0,
      "trained_idea_rejected_candidates": 55,
      "trained_idea_transition_steps": 1,
      "predicted_zero_accepted_violations": true,
      "confirmed": true
    },
    "epsilon_calibration_mse": 0.02,
    "custom_track": null
  },
  "protocol": {
    "paired_seeds": [
      0,
      1,
      2,
      3,
      4,
      5,
      6,
      7
    ],
    "baseline_grid": [
      {
        "lr": 0.0015
      },
      {
        "lr": 0.003
      },
      {
        "lr": 0.006
      }
    ],
    "idea_grid": [
      {
        "lr": 0.0015
      },
      {
        "lr": 0.003
      },
      {
        "lr": 0.006
      }
    ],
    "epochs": 18,
    "n_train": 400,
    "n_test": 200
  }
}