{
  "bench_version": 1,
  "track": "dynamics",
  "model": "rnn_small",
  "metric_direction": "lower is better",
  "n_seeds": 8,
  "baseline": {
    "best_cfg": {
      "lr": 0.01,
      "mu": 0.0
    },
    "sweep": [
      {
        "cfg": {
          "lr": 0.001,
          "mu": 0.0
        },
        "mean": 0.004298779065720737
      },
      {
        "cfg": {
          "lr": 0.001,
          "mu": 0.05
        },
        "mean": 0.004298779065720737
      },
      {
        "cfg": {
          "lr": 0.001,
          "mu": 0.2
        },
        "mean": 0.004298779065720737
      },
      {
        "cfg": {
          "lr": 0.001,
          "mu": 0.5
        },
        "mean": 0.004298779065720737
      },
      {
        "cfg": {
          "lr": 0.003,
          "mu": 0.0
        },
        "mean": 0.0034615940821822733
      },
      {
        "cfg": {
          "lr": 0.003,
          "mu": 0.05
        },
        "mean": 0.0034615940821822733
      },
      {
        "cfg": {
          "lr": 0.003,
          "mu": 0.2
        },
        "mean": 0.0034615940821822733
      },
      {
        "cfg": {
          "lr": 0.003,
          "mu": 0.5
        },
        "mean": 0.0034615940821822733
      },
      {
        "cfg": {
          "lr": 0.01,
          "mu": 0.0
        },
        "mean": 0.0014865802077110857
      },
      {
        "cfg": {
          "lr": 0.01,
          "mu": 0.05
        },
        "mean": 0.0014865802077110857
      },
      {
        "cfg": {
          "lr": 0.01,
          "mu": 0.2
        },
        "mean": 0.0014865802077110857
      },
      {
        "cfg": {
          "lr": 0.01,
          "mu": 0.5
        },
        "mean": 0.0014865802077110857
      }
    ],
    "full": {
      "mean": 0.0014198302378645167,
      "std": 0.00039839114901047844,
      "per_seed": [
        0.0009327387670055032,
        0.0018641706556081772,
        0.0010785290505737066,
        0.0020708823576569557,
        0.0009260809747502208,
        0.0014055331703275442,
        0.0016443432541564107,
        0.001436363672837615
      ],
      "n": 8
    }
  },
  "idea": {
    "mean": 0.0013042011414654553,
    "std": 0.0005283693045785661,
    "per_seed": [
      0.0023362392093986273,
      0.0014533286448568106,
      0.0007090693688951433,
      0.0012600651243701577,
      0.0010053017176687717,
      0.001083051203750074,
      0.0007225603912957013,
      0.0018639934714883566
    ],
    "n": 8,
    "best_cfg": {
      "lr": 0.01,
      "mu": 0.05
    },
    "sweep": [
      {
        "cfg": {
          "lr": 0.001,
          "mu": 0.05
        },
        "mean": 0.004056878213305026
      },
      {
        "cfg": {
          "lr": 0.001,
          "mu": 0.2
        },
        "mean": 0.004066162509843707
      },
      {
        "cfg": {
          "lr": 0.001,
          "mu": 0.5
        },
        "mean": 0.004089910595212132
      },
      {
        "cfg": {
          "lr": 0.003,
          "mu": 0.05
        },
        "mean": 0.003846683946903795
      },
      {
        "cfg": {
          "lr": 0.003,
          "mu": 0.2
        },
        "mean": 0.003857339732348919
      },
      {
        "cfg": {
          "lr": 0.003,
          "mu": 0.5
        },
        "mean": 0.0038832707796245813
      },
      {
        "cfg": {
          "lr": 0.01,
          "mu": 0.05
        },
        "mean": 0.0014396755868801847
      },
      {
        "cfg": {
          "lr": 0.01,
          "mu": 0.2
        },
        "mean": 0.0014551622880389914
      },
      {
        "cfg": {
          "lr": 0.01,
          "mu": 0.5
        },
        "mean": 0.001473813972552307
      }
    ]
  },
  "comparison": {
    "delta_mean": -0.00011562909639906138,
    "idea_wins": 5,
    "n_pairs": 8,
    "per_seed_diffs": [
      0.001403500442393124,
      -0.0004108420107513666,
      -0.00036945968167856336,
      -0.000810817233286798,
      7.922074291855097e-05,
      -0.0003224819665774703,
      -0.0009217828628607094,
      0.0004276297986507416
    ],
    "p_value": 0.6789,
    "mde": 0.0006295704392970048,
    "mde_rel_pct": 44.341247460957355,
    "verdict": "no significant win",
    "system_worked": false
  },
  "track_justification": "Dynamics is structurally matched: the task is controlled pendulum rollout and the intervention constrains predicted outputs under empirical error scenarios.",
  "idea_sweep": [
    {
      "cfg": {
        "lr": 0.001,
        "mu": 0.05
      },
      "mean": 0.004056878213305026
    },
    {
      "cfg": {
        "lr": 0.001,
        "mu": 0.2
      },
      "mean": 0.004066162509843707
    },
    {
      "cfg": {
        "lr": 0.001,
        "mu": 0.5
      },
      "mean": 0.004089910595212132
    },
    {
      "cfg": {
        "lr": 0.003,
        "mu": 0.05
      },
      "mean": 0.003846683946903795
    },
    {
      "cfg": {
        "lr": 0.003,
        "mu": 0.2
      },
      "mean": 0.003857339732348919
    },
    {
      "cfg": {
        "lr": 0.003,
        "mu": 0.5
      },
      "mean": 0.0038832707796245813
    },
    {
      "cfg": {
        "lr": 0.01,
        "mu": 0.05
      },
      "mean": 0.0014396755868801847
    },
    {
      "cfg": {
        "lr": 0.01,
        "mu": 0.2
      },
      "mean": 0.0014551622880389914
    },
    {
      "cfg": {
        "lr": 0.01,
        "mu": 0.5
      },
      "mean": 0.001473813972552307
    }
  ],
  "custom_track": null,
  "mechanism_signature": {
    "baseline_residual_std": 0.034058332443237305,
    "idea_residual_std": 0.031945905182510614,
    "baseline_mean_slack": 0.0,
    "idea_mean_slack": 0.0,
    "prediction": "scenario penalty reduces empirical residual-boundary slack",
    "confirmed": false
  }
}