{
  "format": "vla-moe-public-progress-v1",
  "asOf": "2026-10-07T10:53:03.291Z",
  "displayTime": "2026-10-07 18:53",
  "timeZone": "Asia/Shanghai",
  "mode": "static_snapshot",
  "protocol": {
    "effectiveBatch": 64,
    "microbatch": 4,
    "gradientAccumulation": 16,
    "optimizerUpdates": 9375,
    "evalSeed": 42,
    "episodesPerSuite": 500,
    "primaryRouting": "oracle_suite_id"
  },
  "primary": [
    {
      "method": "Rank8",
      "seed": 1,
      "environment": "AI088",
      "counts": [
        490,
        499,
        490,
        462
      ],
      "episodes": [
        500,
        500,
        500,
        500
      ],
      "suiteRates": [
        98,
        99.8,
        98,
        92.4
      ],
      "macro": 97.05,
      "successes": 1941,
      "totalEpisodes": 2000
    },
    {
      "method": "Rank8",
      "seed": 2,
      "environment": "AI088",
      "counts": [
        481,
        493,
        487,
        439
      ],
      "episodes": [
        500,
        500,
        500,
        500
      ],
      "suiteRates": [
        96.2,
        98.6,
        97.4,
        87.8
      ],
      "macro": 95,
      "successes": 1900,
      "totalEpisodes": 2000
    },
    {
      "method": "Rank8",
      "seed": 3,
      "environment": "AI088",
      "counts": [
        487,
        498,
        487,
        475
      ],
      "episodes": [
        500,
        500,
        500,
        500
      ],
      "suiteRates": [
        97.4,
        99.6,
        97.4,
        95
      ],
      "macro": 97.35,
      "successes": 1947,
      "totalEpisodes": 2000
    },
    {
      "method": "Rank8",
      "seed": 4,
      "environment": "AI088",
      "counts": [
        495,
        496,
        497,
        476
      ],
      "episodes": [
        500,
        500,
        500,
        500
      ],
      "suiteRates": [
        99,
        99.2,
        99.4,
        95.2
      ],
      "macro": 98.2,
      "successes": 1964,
      "totalEpisodes": 2000
    },
    {
      "method": "Dense",
      "seed": 1,
      "environment": "AI088",
      "counts": [
        488,
        497,
        491,
        463
      ],
      "episodes": [
        500,
        500,
        500,
        500
      ],
      "suiteRates": [
        97.6,
        99.4,
        98.2,
        92.6
      ],
      "macro": 96.95,
      "successes": 1939,
      "totalEpisodes": 2000
    },
    {
      "method": "Dense",
      "seed": 2,
      "environment": "AI088",
      "counts": [
        477,
        497,
        480,
        465
      ],
      "episodes": [
        500,
        500,
        500,
        500
      ],
      "suiteRates": [
        95.4,
        99.4,
        96,
        93
      ],
      "macro": 95.95,
      "successes": 1919,
      "totalEpisodes": 2000
    },
    {
      "method": "Dense",
      "seed": 3,
      "environment": "AI088",
      "counts": [
        487,
        484,
        384,
        426
      ],
      "episodes": [
        500,
        500,
        500,
        500
      ],
      "suiteRates": [
        97.4,
        96.8,
        76.8,
        85.2
      ],
      "macro": 89.05,
      "successes": 1781,
      "totalEpisodes": 2000
    },
    {
      "method": "Dense",
      "seed": 4,
      "environment": "AI088",
      "counts": [
        490,
        500,
        469,
        465
      ],
      "episodes": [
        500,
        500,
        500,
        500
      ],
      "suiteRates": [
        98,
        100,
        93.8,
        93
      ],
      "macro": 96.2,
      "successes": 1924,
      "totalEpisodes": 2000
    }
  ],
  "statistics": {
    "rank8": {
      "seeds": [
        1,
        2,
        3,
        4
      ],
      "suiteMean": [
        97.65,
        99.3,
        98.05,
        92.6
      ],
      "suiteSD": [
        1.17046999,
        0.52915026,
        0.94339811,
        3.44480285
      ],
      "macroMean": 96.9,
      "macroSD": 1.3570802,
      "successes": 7752,
      "episodes": 8000
    },
    "dense": {
      "seeds": [
        1,
        2,
        3,
        4
      ],
      "suiteMean": [
        97.1,
        98.9,
        91.2,
        90.95
      ],
      "suiteSD": [
        1.16045968,
        1.42828569,
        9.76660978,
        3.83796821
      ],
      "macroMean": 94.5375,
      "macroSD": 3.68292796,
      "successes": 7563,
      "episodes": 8000
    },
    "delta": 2.3625,
    "paired": [
      {
        "seed": 1,
        "delta": 0.1
      },
      {
        "seed": 2,
        "delta": -0.95
      },
      {
        "seed": 3,
        "delta": 8.3
      },
      {
        "seed": 4,
        "delta": 2
      }
    ]
  },
  "rank7": [
    {
      "method": "Rank7",
      "seed": 2,
      "environment": "AI088",
      "counts": [
        486,
        497,
        494,
        449
      ],
      "episodes": [
        500,
        500,
        500,
        500
      ],
      "suiteRates": [
        97.2,
        99.4,
        98.8,
        89.8
      ],
      "macro": 96.3,
      "successes": 1926,
      "totalEpisodes": 2000
    },
    {
      "method": "Rank7",
      "seed": 4,
      "environment": "AI088",
      "counts": [
        493,
        499,
        383,
        463
      ],
      "episodes": [
        500,
        500,
        500,
        500
      ],
      "suiteRates": [
        98.6,
        99.8,
        76.6,
        92.6
      ],
      "macro": 91.9,
      "successes": 1838,
      "totalEpisodes": 2000
    }
  ],
  "crossServerDense": [
    {
      "method": "Dense",
      "seed": 1,
      "environment": "AI081",
      "counts": [
        493,
        499,
        491,
        460
      ],
      "episodes": [
        500,
        500,
        500,
        500
      ],
      "suiteRates": [
        98.6,
        99.8,
        98.2,
        92
      ],
      "macro": 97.15,
      "successes": 1943,
      "totalEpisodes": 2000
    },
    {
      "method": "Dense",
      "seed": 2,
      "environment": "AI081",
      "counts": [
        430,
        500,
        138,
        413
      ],
      "episodes": [
        500,
        500,
        500,
        500
      ],
      "suiteRates": [
        86,
        100,
        27.6,
        82.6
      ],
      "macro": 74.05,
      "successes": 1481,
      "totalEpisodes": 2000
    },
    {
      "method": "Dense",
      "seed": 3,
      "environment": "AI081",
      "counts": [
        487,
        496,
        400,
        432
      ],
      "episodes": [
        500,
        500,
        500,
        500
      ],
      "suiteRates": [
        97.4,
        99.2,
        80,
        86.4
      ],
      "macro": 90.75,
      "successes": 1815,
      "totalEpisodes": 2000
    },
    {
      "method": "Dense",
      "seed": 4,
      "environment": "AI081",
      "counts": [
        488,
        497,
        403,
        455
      ],
      "episodes": [
        500,
        500,
        500,
        500
      ],
      "suiteRates": [
        97.6,
        99.4,
        80.6,
        91
      ],
      "macro": 92.15,
      "successes": 1843,
      "totalEpisodes": 2000
    }
  ],
  "crossServerStatistics": {
    "macroMean": 88.525,
    "macroSD": 10.03340255
  },
  "progress": [
    {
      "id": "ai081-v15-seed3",
      "method": "V15 Tail-Residual Rank8",
      "seed": 3,
      "environment": "AI081",
      "sourceGeneratedAt": "2026-10-07T10:52:23.524249+00:00",
      "training": {
        "status": "complete",
        "step": 9375,
        "target": 9375
      },
      "evaluation": {
        "status": "running",
        "episodes": 1497,
        "target": 2000,
        "completedSuites": 2,
        "finalMacro": null,
        "suites": [
          {
            "suite": "spatial",
            "status": "complete",
            "episodes": 500,
            "target": 500,
            "finalSuccesses": 496,
            "finalRate": 99.2
          },
          {
            "suite": "object",
            "status": "complete",
            "episodes": 500,
            "target": 500,
            "finalSuccesses": 495,
            "finalRate": 99
          },
          {
            "suite": "goal",
            "status": "running",
            "episodes": 279,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          },
          {
            "suite": "long",
            "status": "running",
            "episodes": 218,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          }
        ]
      },
      "repeat": false
    },
    {
      "id": "ai088-dense-seed3-repeat1",
      "method": "Official Dense",
      "seed": 3,
      "environment": "AI088",
      "sourceGeneratedAt": "2026-10-07T10:52:16.189785+00:00",
      "training": {
        "status": "running",
        "step": 6413,
        "target": 9375
      },
      "evaluation": {
        "status": "not-started",
        "episodes": 0,
        "target": 2000,
        "completedSuites": 0,
        "finalMacro": null,
        "suites": [
          {
            "suite": "spatial",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          },
          {
            "suite": "object",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          },
          {
            "suite": "goal",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          },
          {
            "suite": "long",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          }
        ]
      },
      "repeat": true
    },
    {
      "id": "ai088-v15-seed2-repeat1",
      "method": "V15 Tail-Residual Rank8",
      "seed": 2,
      "environment": "AI088",
      "sourceGeneratedAt": "2026-10-07T10:51:35.066492+00:00",
      "training": {
        "status": "running",
        "step": 3855,
        "target": 9375
      },
      "evaluation": {
        "status": "not-started",
        "episodes": 0,
        "target": 2000,
        "completedSuites": 0,
        "finalMacro": null,
        "suites": [
          {
            "suite": "spatial",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          },
          {
            "suite": "object",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          },
          {
            "suite": "goal",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          },
          {
            "suite": "long",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          }
        ]
      },
      "repeat": true
    },
    {
      "id": "ai088-affine-seed3",
      "method": "V15 Output-Affine Calibration",
      "seed": 3,
      "environment": "AI088",
      "sourceGeneratedAt": "2026-10-07T10:51:56.120146+00:00",
      "training": {
        "status": "queued",
        "step": 0,
        "target": 9375
      },
      "evaluation": {
        "status": "not-started",
        "episodes": 0,
        "target": 2000,
        "completedSuites": 0,
        "finalMacro": null,
        "suites": [
          {
            "suite": "spatial",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          },
          {
            "suite": "object",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          },
          {
            "suite": "goal",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          },
          {
            "suite": "long",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          }
        ]
      },
      "repeat": false
    },
    {
      "id": "ai088-affine-seed1",
      "method": "V15 Output-Affine Calibration",
      "seed": 1,
      "environment": "AI088",
      "sourceGeneratedAt": "2026-10-07T10:51:59.165705+00:00",
      "training": {
        "status": "queued",
        "step": 0,
        "target": 9375
      },
      "evaluation": {
        "status": "not-started",
        "episodes": 0,
        "target": 2000,
        "completedSuites": 0,
        "finalMacro": null,
        "suites": [
          {
            "suite": "spatial",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          },
          {
            "suite": "object",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          },
          {
            "suite": "goal",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          },
          {
            "suite": "long",
            "status": "waiting",
            "episodes": 0,
            "target": 500,
            "finalSuccesses": null,
            "finalRate": null
          }
        ]
      },
      "repeat": false
    }
  ],
  "sources": [
    {
      "name": "Rank8 / Dense final audit",
      "date": "2026-10-04"
    },
    {
      "name": "Cross-server Dense final audit",
      "date": "2026-10-04"
    },
    {
      "name": "Read-only experiment monitoring snapshot",
      "collectedAt": "2026-10-07T10:53:03.291Z"
    }
  ],
  "notes": [
    "Four training seeds per primary method; sample SD in percentage points.",
    "Technical repeats and cross-server diagnostics are not additional independent seeds.",
    "Incomplete evaluations have no final Macro and are excluded from primary means.",
    "The current mean difference is descriptive; statistical significance, generalization, and causal mechanisms are not established."
  ]
}
