{
  "schemaVersion": 1,
  "studyId": "leduc-reward-exploratory-scaled-v1",
  "sourceStudyId": "leduc-reward-study-v1",
  "analysisType": "post_outcome_exploratory",
  "confirmatory": false,
  "generatedAt": "2026-08-10T17:12:16.048769+00:00",
  "arms": [
    "terminal",
    "scaled"
  ],
  "seeds": [
    31415,
    27182,
    16180,
    57721,
    65537,
    99991,
    104729,
    13007,
    17011,
    23003
  ],
  "completedRuns": 20,
  "totalEpisodes": 6000000,
  "metrics": {
    "finalExploitability": {
      "differenceDirection": "scaled_minus_terminal",
      "perSeed": [
        {
          "seed": 31415,
          "terminal": 1.1656108377862684,
          "scaled": 1.3103060408157416,
          "difference": 0.14469520302947325
        },
        {
          "seed": 27182,
          "terminal": 1.2028935981706241,
          "scaled": 1.3044224516261569,
          "difference": 0.10152885345553275
        },
        {
          "seed": 16180,
          "terminal": 1.1080523367981727,
          "scaled": 1.3310925804591314,
          "difference": 0.22304024366095865
        },
        {
          "seed": 57721,
          "terminal": 1.2129657857157117,
          "scaled": 1.340850943559785,
          "difference": 0.1278851578440734
        },
        {
          "seed": 65537,
          "terminal": 1.1505005941537312,
          "scaled": 1.3173182505961791,
          "difference": 0.16681765644244795
        },
        {
          "seed": 99991,
          "terminal": 1.1841468647174214,
          "scaled": 1.201313987042301,
          "difference": 0.017167122324879713
        },
        {
          "seed": 104729,
          "terminal": 1.1956021309598137,
          "scaled": 1.2968633614914904,
          "difference": 0.10126123053167668
        },
        {
          "seed": 13007,
          "terminal": 1.2341049464708413,
          "scaled": 1.2417504468758553,
          "difference": 0.007645500405013994
        },
        {
          "seed": 17011,
          "terminal": 1.1564215825051036,
          "scaled": 1.3263073990155114,
          "difference": 0.1698858165104078
        },
        {
          "seed": 23003,
          "terminal": 1.1855280481569292,
          "scaled": 1.3049482439792879,
          "difference": 0.11942019582235863
        }
      ],
      "meanDifference": 0.11793469800268228,
      "medianDifference": 0.12365267683321601,
      "bootstrap95CI": [
        0.0780670554505237,
        0.15537807473427334
      ],
      "exactPairedSignFlipP": 0.001953125,
      "pairedStandardizedMeanDifference": 1.7730599193022425,
      "interpretation": "exploratory_only"
    },
    "exploitabilityAuc": {
      "differenceDirection": "scaled_minus_terminal",
      "perSeed": [
        {
          "seed": 31415,
          "terminal": 320948.92802255444,
          "scaled": 363473.08808323304,
          "difference": 42524.16006067861
        },
        {
          "seed": 27182,
          "terminal": 322606.57886433817,
          "scaled": 352227.506777913,
          "difference": 29620.92791357485
        },
        {
          "seed": 16180,
          "terminal": 319434.76784389856,
          "scaled": 364068.7991399553,
          "difference": 44634.03129605675
        },
        {
          "seed": 57721,
          "terminal": 375440.1806374968,
          "scaled": 356980.8375736869,
          "difference": -18459.34306380991
        },
        {
          "seed": 65537,
          "terminal": 318397.75586208294,
          "scaled": 362537.3640983912,
          "difference": 44139.60823630827
        },
        {
          "seed": 99991,
          "terminal": 326358.0541633002,
          "scaled": 337362.9721623257,
          "difference": 11004.917999025492
        },
        {
          "seed": 104729,
          "terminal": 323269.29412799084,
          "scaled": 354697.3133263072,
          "difference": 31428.019198316382
        },
        {
          "seed": 13007,
          "terminal": 330252.79858153214,
          "scaled": 361844.36200238264,
          "difference": 31591.563420850493
        },
        {
          "seed": 17011,
          "terminal": 322193.9247089877,
          "scaled": 376648.41193263006,
          "difference": 54454.487223642354
        },
        {
          "seed": 23003,
          "terminal": 328840.22972364293,
          "scaled": 357400.2049534572,
          "difference": 28559.9752298143
        }
      ],
      "meanDifference": 29949.83475144576,
      "medianDifference": 31509.791309583437,
      "bootstrap95CI": [
        16511.385499777112,
        40581.75695005607
      ],
      "exactPairedSignFlipP": 0.005859375,
      "pairedStandardizedMeanDifference": 1.4415470076664798,
      "interpretation": "exploratory_only",
      "window": [
        25000,
        300000
      ]
    }
  },
  "exploratoryRationale": {
    "arm": "scaled",
    "basis": "Lowest Pilot exploitability AUC among non-control arms and the simplest reward transformation.",
    "pilotSupport": false
  },
  "claimLimit": "Post-outcome exploratory evidence only; not a confirmation and not a retroactive Pilot promotion."
}