{
  "bench_version": 1,
  "track": "phase_checkpoint_regression",
  "model": "mlp_tiny",
  "metric_direction": "lower is better",
  "n_seeds": 8,
  "baseline": {
    "best_cfg": {
      "lr": 0.01,
      "weight_decay": 0.001,
      "epochs": 12,
      "sigma": 0.05
    },
    "sweep": [
      {
        "cfg": {
          "lr": 0.001,
          "weight_decay": 0.0,
          "epochs": 12,
          "sigma": 0.0
        },
        "mean": 0.052177092991769314
      },
      {
        "cfg": {
          "lr": 0.001,
          "weight_decay": 0.0001,
          "epochs": 12,
          "sigma": 0.02
        },
        "mean": 0.05208037327975035
      },
      {
        "cfg": {
          "lr": 0.001,
          "weight_decay": 0.001,
          "epochs": 12,
          "sigma": 0.05
        },
        "mean": 0.051703399047255516
      },
      {
        "cfg": {
          "lr": 0.003,
          "weight_decay": 0.0,
          "epochs": 12,
          "sigma": 0.0
        },
        "mean": 0.03177726315334439
      },
      {
        "cfg": {
          "lr": 0.003,
          "weight_decay": 0.0001,
          "epochs": 12,
          "sigma": 0.02
        },
        "mean": 0.03205729741603136
      },
      {
        "cfg": {
          "lr": 0.003,
          "weight_decay": 0.001,
          "epochs": 12,
          "sigma": 0.05
        },
        "mean": 0.03150508273392916
      },
      {
        "cfg": {
          "lr": 0.01,
          "weight_decay": 0.0,
          "epochs": 12,
          "sigma": 0.0
        },
        "mean": 0.022574170026928186
      },
      {
        "cfg": {
          "lr": 0.01,
          "weight_decay": 0.0001,
          "epochs": 12,
          "sigma": 0.02
        },
        "mean": 0.022592097986489534
      },
      {
        "cfg": {
          "lr": 0.01,
          "weight_decay": 0.001,
          "epochs": 12,
          "sigma": 0.05
        },
        "mean": 0.02199754724279046
      }
    ],
    "full": {
      "mean": 0.022857969626784325,
      "std": 0.0019890920183647087,
      "per_seed": [
        0.022877627983689308,
        0.02057631127536297,
        0.02327309176325798,
        0.021263157948851585,
        0.0209011472761631,
        0.026269573718309402,
        0.02207833342254162,
        0.025624513626098633
      ],
      "n": 8
    }
  },
  "idea": {
    "mean": 0.022857969626784325,
    "std": 0.0019890920183647087,
    "per_seed": [
      0.022877627983689308,
      0.02057631127536297,
      0.02327309176325798,
      0.021263157948851585,
      0.0209011472761631,
      0.026269573718309402,
      0.02207833342254162,
      0.025624513626098633
    ],
    "n": 8
  },
  "comparison": {
    "delta_mean": 0.0,
    "idea_wins": 0,
    "n_pairs": 8,
    "per_seed_diffs": [
      0.0,
      0.0,
      0.0,
      0.0,
      0.0,
      0.0,
      0.0,
      0.0
    ],
    "p_value": 1.0,
    "mde": 0.0,
    "mde_rel_pct": 0.0,
    "verdict": "no measurable effect",
    "system_worked": false
  },
  "mechanism_signature": {
    "prediction": "anonymous trained-system checkpoint phase return slope = 1",
    "predicted_slope": 1.0,
    "observed_slope_mean": 0.04611545590999941,
    "priority_observed_slope_mean": 1.0049859781212493,
    "observed_gap_drift_mean": 2.667137344314341e-18,
    "confirmed": false,
    "custom_track": {
      "name": "phase_checkpoint_regression",
      "file": "phase_track.py",
      "domain": "training-dynamics"
    },
    "idea_sweep": [
      {
        "cfg": {
          "lr": 0.01,
          "weight_decay": 0.001,
          "epochs": 12,
          "sigma": 0.05
        },
        "mean": 0.02199754724279046
      },
      {
        "cfg": {
          "lr": 0.001,
          "weight_decay": 0.001,
          "epochs": 12,
          "sigma": 0.05
        },
        "mean": 0.051703399047255516
      },
      {
        "cfg": {
          "lr": 0.01,
          "weight_decay": 0.001,
          "epochs": 12,
          "sigma": 0.05
        },
        "mean": 0.02199754724279046
      }
    ],
    "selected_idea_cfg": {
      "lr": 0.01,
      "weight_decay": 0.001,
      "epochs": 12,
      "sigma": 0.05
    }
  }
}