{
  "project": "EvalFlow CI Lab",
  "type": "fixed-snapshot-research-gate",
  "status": "PASS",
  "baseline": "nb",
  "candidate": "hybrid",
  "split": "challenge",
  "policy": {
    "minN": 50,
    "minAccuracy": 0.75,
    "minMacroF1": 0.72,
    "maxRegression": 0.03,
    "maxTypoFlip": 0.05,
    "maxContextFlip": 0.15,
    "minLanguageN": 15,
    "minLanguageAccuracy": 0.65,
    "requireComplete": true
  },
  "baselineMetrics": {
    "method": "nb",
    "split": "challenge",
    "n": 54,
    "accuracy": 0.7407407407407407,
    "macroF1": 0.7362809928289805,
    "classes": [
      {
        "label": "background",
        "n": 9,
        "tp": 4,
        "fp": 0,
        "fn": 5,
        "precision": 1,
        "recall": 0.4444444444444444,
        "f1": 0.6153846153846153
      },
      {
        "label": "objective",
        "n": 9,
        "tp": 7,
        "fp": 1,
        "fn": 2,
        "precision": 0.875,
        "recall": 0.7777777777777778,
        "f1": 0.823529411764706
      },
      {
        "label": "method",
        "n": 9,
        "tp": 8,
        "fp": 2,
        "fn": 1,
        "precision": 0.8,
        "recall": 0.8888888888888888,
        "f1": 0.8421052631578948
      },
      {
        "label": "result",
        "n": 9,
        "tp": 6,
        "fp": 1,
        "fn": 3,
        "precision": 0.8571428571428571,
        "recall": 0.6666666666666666,
        "f1": 0.75
      },
      {
        "label": "limitation",
        "n": 9,
        "tp": 9,
        "fp": 7,
        "fn": 0,
        "precision": 0.5625,
        "recall": 1,
        "f1": 0.72
      },
      {
        "label": "implication",
        "n": 9,
        "tp": 6,
        "fp": 3,
        "fn": 3,
        "precision": 0.6666666666666666,
        "recall": 0.6666666666666666,
        "f1": 0.6666666666666666
      }
    ],
    "groups": {
      "language": [
        {
          "label": "en",
          "n": 18,
          "accuracy": 0.8333333333333334,
          "errors": 3
        },
        {
          "label": "es",
          "n": 18,
          "accuracy": 0.6111111111111112,
          "errors": 7
        },
        {
          "label": "pt",
          "n": 18,
          "accuracy": 0.7777777777777778,
          "errors": 4
        }
      ],
      "domain": [
        {
          "label": "mixed",
          "n": 54,
          "accuracy": 0.7407407407407407,
          "errors": 14
        }
      ],
      "truth": [
        {
          "label": "background",
          "n": 9,
          "accuracy": 0.4444444444444444,
          "errors": 5
        },
        {
          "label": "implication",
          "n": 9,
          "accuracy": 0.6666666666666666,
          "errors": 3
        },
        {
          "label": "limitation",
          "n": 9,
          "accuracy": 1,
          "errors": 0
        },
        {
          "label": "method",
          "n": 9,
          "accuracy": 0.8888888888888888,
          "errors": 1
        },
        {
          "label": "objective",
          "n": 9,
          "accuracy": 0.7777777777777778,
          "errors": 2
        },
        {
          "label": "result",
          "n": 9,
          "accuracy": 0.6666666666666666,
          "errors": 3
        }
      ]
    },
    "attacks": [
      {
        "id": "original",
        "n": 54,
        "matched": 0,
        "coverage": 1,
        "changedShare": 0,
        "flipRate": 0,
        "flipN": 0,
        "matchedAccuracy": null,
        "matchedOriginalAccuracy": null,
        "destructive": false
      },
      {
        "id": "format",
        "n": 54,
        "matched": 54,
        "coverage": 1,
        "changedShare": 1,
        "flipRate": 0,
        "flipN": 0,
        "matchedAccuracy": 0.7407407407407407,
        "matchedOriginalAccuracy": 0.7407407407407407,
        "destructive": false
      },
      {
        "id": "typo",
        "n": 54,
        "matched": 54,
        "coverage": 1,
        "changedShare": 1,
        "flipRate": 0.05555555555555555,
        "flipN": 3,
        "matchedAccuracy": 0.6851851851851852,
        "matchedOriginalAccuracy": 0.7407407407407407,
        "destructive": false
      },
      {
        "id": "synonym",
        "n": 54,
        "matched": 9,
        "coverage": 1,
        "changedShare": 0.16666666666666666,
        "flipRate": 0,
        "flipN": 0,
        "matchedAccuracy": 0.8888888888888888,
        "matchedOriginalAccuracy": 0.8888888888888888,
        "destructive": false
      },
      {
        "id": "context",
        "n": 54,
        "matched": 54,
        "coverage": 1,
        "changedShare": 1,
        "flipRate": 0.2962962962962963,
        "flipN": 16,
        "matchedAccuracy": 0.6296296296296297,
        "matchedOriginalAccuracy": 0.7407407407407407,
        "destructive": false
      },
      {
        "id": "truncate",
        "n": 54,
        "matched": 54,
        "coverage": 1,
        "changedShare": 1,
        "flipRate": 0.12962962962962962,
        "flipN": 7,
        "matchedAccuracy": null,
        "matchedOriginalAccuracy": null,
        "destructive": true
      }
    ]
  },
  "candidateMetrics": {
    "method": "hybrid",
    "split": "challenge",
    "n": 54,
    "accuracy": 0.7777777777777778,
    "macroF1": 0.7617571707045391,
    "classes": [
      {
        "label": "background",
        "n": 9,
        "tp": 3,
        "fp": 0,
        "fn": 6,
        "precision": 1,
        "recall": 0.3333333333333333,
        "f1": 0.5
      },
      {
        "label": "objective",
        "n": 9,
        "tp": 8,
        "fp": 1,
        "fn": 1,
        "precision": 0.8888888888888888,
        "recall": 0.8888888888888888,
        "f1": 0.8888888888888888
      },
      {
        "label": "method",
        "n": 9,
        "tp": 9,
        "fp": 1,
        "fn": 0,
        "precision": 0.9,
        "recall": 1,
        "f1": 0.9473684210526316
      },
      {
        "label": "result",
        "n": 9,
        "tp": 5,
        "fp": 0,
        "fn": 4,
        "precision": 1,
        "recall": 0.5555555555555556,
        "f1": 0.7142857142857143
      },
      {
        "label": "limitation",
        "n": 9,
        "tp": 9,
        "fp": 7,
        "fn": 0,
        "precision": 0.5625,
        "recall": 1,
        "f1": 0.72
      },
      {
        "label": "implication",
        "n": 9,
        "tp": 8,
        "fp": 3,
        "fn": 1,
        "precision": 0.7272727272727273,
        "recall": 0.8888888888888888,
        "f1": 0.7999999999999999
      }
    ],
    "groups": {
      "language": [
        {
          "label": "en",
          "n": 18,
          "accuracy": 0.7777777777777778,
          "errors": 4
        },
        {
          "label": "es",
          "n": 18,
          "accuracy": 0.7222222222222222,
          "errors": 5
        },
        {
          "label": "pt",
          "n": 18,
          "accuracy": 0.8333333333333334,
          "errors": 3
        }
      ],
      "domain": [
        {
          "label": "mixed",
          "n": 54,
          "accuracy": 0.7777777777777778,
          "errors": 12
        }
      ],
      "truth": [
        {
          "label": "background",
          "n": 9,
          "accuracy": 0.3333333333333333,
          "errors": 6
        },
        {
          "label": "implication",
          "n": 9,
          "accuracy": 0.8888888888888888,
          "errors": 1
        },
        {
          "label": "limitation",
          "n": 9,
          "accuracy": 1,
          "errors": 0
        },
        {
          "label": "method",
          "n": 9,
          "accuracy": 1,
          "errors": 0
        },
        {
          "label": "objective",
          "n": 9,
          "accuracy": 0.8888888888888888,
          "errors": 1
        },
        {
          "label": "result",
          "n": 9,
          "accuracy": 0.5555555555555556,
          "errors": 4
        }
      ]
    },
    "attacks": [
      {
        "id": "original",
        "n": 54,
        "matched": 0,
        "coverage": 1,
        "changedShare": 0,
        "flipRate": 0,
        "flipN": 0,
        "matchedAccuracy": null,
        "matchedOriginalAccuracy": null,
        "destructive": false
      },
      {
        "id": "format",
        "n": 54,
        "matched": 54,
        "coverage": 1,
        "changedShare": 1,
        "flipRate": 0,
        "flipN": 0,
        "matchedAccuracy": 0.7777777777777778,
        "matchedOriginalAccuracy": 0.7777777777777778,
        "destructive": false
      },
      {
        "id": "typo",
        "n": 54,
        "matched": 54,
        "coverage": 1,
        "changedShare": 1,
        "flipRate": 0.018518518518518517,
        "flipN": 1,
        "matchedAccuracy": 0.7592592592592593,
        "matchedOriginalAccuracy": 0.7777777777777778,
        "destructive": false
      },
      {
        "id": "synonym",
        "n": 54,
        "matched": 9,
        "coverage": 1,
        "changedShare": 0.16666666666666666,
        "flipRate": 0.018518518518518517,
        "flipN": 1,
        "matchedAccuracy": 1,
        "matchedOriginalAccuracy": 0.8888888888888888,
        "destructive": false
      },
      {
        "id": "context",
        "n": 54,
        "matched": 54,
        "coverage": 1,
        "changedShare": 1,
        "flipRate": 0.09259259259259259,
        "flipN": 5,
        "matchedAccuracy": 0.7222222222222222,
        "matchedOriginalAccuracy": 0.7777777777777778,
        "destructive": false
      },
      {
        "id": "truncate",
        "n": 54,
        "matched": 54,
        "coverage": 1,
        "changedShare": 1,
        "flipRate": 0.1111111111111111,
        "flipN": 6,
        "matchedAccuracy": null,
        "matchedOriginalAccuracy": null,
        "destructive": true
      }
    ]
  },
  "pair": {
    "n": 54,
    "delta": 0.037037037037037035,
    "improved": 4,
    "regressed": 2,
    "onlyBase": [],
    "onlyCandidate": []
  },
  "bootstrap": {
    "point": 0.037037037037037035,
    "low": -0.037037037037037035,
    "high": 0.12962962962962962,
    "iterations": 500,
    "seed": 241,
    "n": 54,
    "assumptions": "Paired example bootstrap with fixed model predictions; excludes retraining and data-label uncertainty."
  },
  "gates": [
    {
      "id": "n",
      "label": "Minimum held-out examples",
      "actual": 54,
      "limit": 50,
      "comparator": "≥",
      "pass": true,
      "unit": "count"
    },
    {
      "id": "accuracy",
      "label": "Candidate Accuracy",
      "actual": 0.7777777777777778,
      "limit": 0.75,
      "comparator": "≥",
      "pass": true,
      "unit": "percent"
    },
    {
      "id": "macroF1",
      "label": "Candidate macro-F1",
      "actual": 0.7617571707045391,
      "limit": 0.72,
      "comparator": "≥",
      "pass": true,
      "unit": "percent"
    },
    {
      "id": "regression",
      "label": "Maximum allowed regression",
      "actual": 0.037037037037037035,
      "limit": -0.03,
      "comparator": "≥",
      "pass": true,
      "unit": "delta"
    },
    {
      "id": "typo",
      "label": "Typo prediction-flip rate",
      "actual": 0.018518518518518517,
      "limit": 0.05,
      "comparator": "≤",
      "pass": true,
      "unit": "percent"
    },
    {
      "id": "context",
      "label": "Context prediction-flip rate",
      "actual": 0.09259259259259259,
      "limit": 0.15,
      "comparator": "≤",
      "pass": true,
      "unit": "percent"
    },
    {
      "id": "sliceN",
      "label": "Minimum per-language examples",
      "actual": 18,
      "limit": 15,
      "comparator": "≥",
      "pass": true,
      "unit": "count"
    },
    {
      "id": "sliceAcc",
      "label": "Minimum language Accuracy",
      "actual": 0.7222222222222222,
      "limit": 0.65,
      "comparator": "≥",
      "pass": true,
      "unit": "percent"
    },
    {
      "id": "coverage",
      "label": "Paired evaluation / stress coverage",
      "actual": 54,
      "limit": 54,
      "comparator": "=",
      "pass": true,
      "unit": "count"
    }
  ],
  "limitations": "This is a local research replay, not production CI execution, a trained model update, deployment authorization, or a fairness certification."
}