{
  "winner": "v32",
  "digests": {
    "qwen35-2b": "09de712676e7ed373b82ebfd33bfeb7fa70d00c38a3ff63fe104b89fd8a2334f",
    "gemma4-e2b": "b37049369adfe3d2b653af0ab301a062ca5cbe96ace6aa0d3f2559ee9b563fc2"
  },
  "baseline_reference": "precise",
  "screen": {
    "v01": {
      "n": 8,
      "passed": 0,
      "partial": 3,
      "failed": 5,
      "score": 3,
      "issues": {
        "content": 3,
        "numbers": 2,
        "fidelity": 2,
        "format": 2,
        "display": 1,
        "terms": 1,
        "quote": 1
      },
      "median_ms": 397.19
    },
    "v02": {
      "n": 8,
      "passed": 0,
      "partial": 4,
      "failed": 4,
      "score": 4,
      "issues": {
        "content": 4,
        "format": 4,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 421.62
    },
    "v03": {
      "n": 8,
      "passed": 0,
      "partial": 5,
      "failed": 3,
      "score": 5,
      "issues": {
        "content": 4,
        "terms": 3,
        "format": 3,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 418.87
    },
    "v04": {
      "n": 8,
      "passed": 1,
      "partial": 4,
      "failed": 3,
      "score": 6,
      "issues": {
        "content": 3,
        "format": 2,
        "terms": 2,
        "numbers": 2,
        "display": 1,
        "fidelity": 1,
        "quote": 1
      },
      "median_ms": 366.31
    },
    "v05": {
      "n": 8,
      "passed": 0,
      "partial": 5,
      "failed": 3,
      "score": 5,
      "issues": {
        "content": 4,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "format": 1,
        "quote": 1
      },
      "median_ms": 388.05
    },
    "v06": {
      "n": 8,
      "passed": 0,
      "partial": 5,
      "failed": 3,
      "score": 5,
      "issues": {
        "content": 4,
        "format": 3,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 388.6
    },
    "v07": {
      "n": 8,
      "passed": 0,
      "partial": 4,
      "failed": 4,
      "score": 4,
      "issues": {
        "content": 4,
        "terms": 3,
        "format": 2,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 384.29
    },
    "v08": {
      "n": 8,
      "passed": 0,
      "partial": 5,
      "failed": 3,
      "score": 5,
      "issues": {
        "content": 4,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "format": 1,
        "quote": 1
      },
      "median_ms": 399.83
    },
    "v09": {
      "n": 8,
      "passed": 0,
      "partial": 4,
      "failed": 4,
      "score": 4,
      "issues": {
        "format": 5,
        "content": 4,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 386.52
    },
    "v10": {
      "n": 8,
      "passed": 0,
      "partial": 4,
      "failed": 4,
      "score": 4,
      "issues": {
        "content": 4,
        "terms": 3,
        "format": 2,
        "display": 1,
        "numbers": 1,
        "quote": 1
      },
      "median_ms": 418.58
    },
    "v11": {
      "n": 8,
      "passed": 1,
      "partial": 3,
      "failed": 4,
      "score": 5,
      "issues": {
        "content": 3,
        "format": 2,
        "terms": 2,
        "numbers": 2,
        "display": 1,
        "fidelity": 1,
        "quote": 1
      },
      "median_ms": 397.49
    },
    "v12": {
      "n": 8,
      "passed": 1,
      "partial": 4,
      "failed": 3,
      "score": 6,
      "issues": {
        "content": 3,
        "terms": 2,
        "numbers": 2,
        "display": 1,
        "fidelity": 1,
        "quote": 1
      },
      "median_ms": 298.35
    },
    "v13": {
      "n": 8,
      "passed": 1,
      "partial": 3,
      "failed": 4,
      "score": 5,
      "issues": {
        "content": 3,
        "terms": 2,
        "numbers": 2,
        "display": 1,
        "fidelity": 1,
        "quote": 1
      },
      "median_ms": 371.45
    },
    "v14": {
      "n": 8,
      "passed": 0,
      "partial": 3,
      "failed": 5,
      "score": 3,
      "issues": {
        "content": 4,
        "terms": 2,
        "numbers": 2,
        "display": 1,
        "fidelity": 1,
        "quote": 1
      },
      "median_ms": 368.15
    },
    "v15": {
      "n": 8,
      "passed": 0,
      "partial": 5,
      "failed": 3,
      "score": 5,
      "issues": {
        "content": 4,
        "terms": 2,
        "numbers": 2,
        "display": 1,
        "fidelity": 1,
        "quote": 1,
        "format": 1
      },
      "median_ms": 337.53
    },
    "v16": {
      "n": 8,
      "passed": 1,
      "partial": 3,
      "failed": 4,
      "score": 5,
      "issues": {
        "content": 3,
        "terms": 2,
        "numbers": 2,
        "display": 1,
        "fidelity": 1,
        "quote": 1
      },
      "median_ms": 358.25
    },
    "v17": {
      "n": 8,
      "passed": 1,
      "partial": 3,
      "failed": 4,
      "score": 5,
      "issues": {
        "content": 3,
        "terms": 2,
        "numbers": 2,
        "display": 1,
        "format": 1,
        "fidelity": 1,
        "quote": 1
      },
      "median_ms": 380.89
    },
    "v18": {
      "n": 8,
      "passed": 1,
      "partial": 5,
      "failed": 2,
      "score": 7,
      "issues": {
        "content": 3,
        "terms": 3,
        "format": 2,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 384.88
    },
    "v19": {
      "n": 8,
      "passed": 0,
      "partial": 5,
      "failed": 3,
      "score": 5,
      "issues": {
        "content": 4,
        "terms": 3,
        "format": 2,
        "numbers": 2,
        "display": 1,
        "repetition": 1
      },
      "median_ms": 426.27
    },
    "v20": {
      "n": 8,
      "passed": 1,
      "partial": 4,
      "failed": 3,
      "score": 6,
      "issues": {
        "content": 3,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "format": 1,
        "quote": 1
      },
      "median_ms": 345.95
    },
    "v21": {
      "n": 8,
      "passed": 0,
      "partial": 5,
      "failed": 3,
      "score": 5,
      "issues": {
        "content": 4,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "format": 1,
        "quote": 1
      },
      "median_ms": 379.62
    },
    "v22": {
      "n": 8,
      "passed": 1,
      "partial": 5,
      "failed": 2,
      "score": 7,
      "issues": {
        "content": 3,
        "terms": 3,
        "format": 2,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 420.34
    },
    "v23": {
      "n": 8,
      "passed": 1,
      "partial": 4,
      "failed": 3,
      "score": 6,
      "issues": {
        "content": 3,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "format": 1,
        "quote": 1
      },
      "median_ms": 366.79
    },
    "v24": {
      "n": 8,
      "passed": 0,
      "partial": 5,
      "failed": 3,
      "score": 5,
      "issues": {
        "content": 4,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "format": 1,
        "quote": 1
      },
      "median_ms": 365.96
    },
    "v25": {
      "n": 8,
      "passed": 1,
      "partial": 5,
      "failed": 2,
      "score": 7,
      "issues": {
        "content": 3,
        "terms": 3,
        "format": 2,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 555.97
    },
    "v26": {
      "n": 8,
      "passed": 1,
      "partial": 5,
      "failed": 2,
      "score": 7,
      "issues": {
        "content": 3,
        "terms": 3,
        "format": 2,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 376.32
    },
    "v27": {
      "n": 8,
      "passed": 1,
      "partial": 5,
      "failed": 2,
      "score": 7,
      "issues": {
        "content": 3,
        "terms": 2,
        "numbers": 2,
        "display": 1,
        "punctuation": 1,
        "format": 1,
        "quote": 1
      },
      "median_ms": 430.05
    },
    "v28": {
      "n": 8,
      "passed": 0,
      "partial": 6,
      "failed": 2,
      "score": 6,
      "issues": {
        "content": 3,
        "terms": 3,
        "format": 2,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 450.54
    },
    "v29": {
      "n": 8,
      "passed": 1,
      "partial": 4,
      "failed": 3,
      "score": 6,
      "issues": {
        "content": 3,
        "numbers": 2,
        "terms": 1,
        "repetition": 1,
        "format": 1,
        "fidelity": 1,
        "quote": 1
      },
      "median_ms": 357.47
    },
    "v30": {
      "n": 8,
      "passed": 0,
      "partial": 5,
      "failed": 3,
      "score": 5,
      "issues": {
        "content": 4,
        "format": 3,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 384.02
    },
    "v31": {
      "n": 8,
      "passed": 1,
      "partial": 4,
      "failed": 3,
      "score": 6,
      "issues": {
        "content": 3,
        "format": 2,
        "numbers": 2,
        "display": 1,
        "terms": 1,
        "punctuation": 1,
        "fidelity": 1,
        "quote": 1
      },
      "median_ms": 392.52
    },
    "v32": {
      "n": 8,
      "passed": 0,
      "partial": 7,
      "failed": 1,
      "score": 7,
      "issues": {
        "content": 4,
        "terms": 3,
        "numbers": 2,
        "format": 2,
        "quote": 1
      },
      "median_ms": 457.01
    },
    "v33": {
      "n": 8,
      "passed": 1,
      "partial": 5,
      "failed": 2,
      "score": 7,
      "issues": {
        "content": 3,
        "format": 3,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 401.13
    },
    "v34": {
      "n": 8,
      "passed": 1,
      "partial": 4,
      "failed": 3,
      "score": 6,
      "issues": {
        "content": 3,
        "numbers": 2,
        "terms": 1,
        "punctuation": 1,
        "fidelity": 1,
        "format": 1,
        "quote": 1
      },
      "median_ms": 330.91
    },
    "v35": {
      "n": 8,
      "passed": 1,
      "partial": 5,
      "failed": 2,
      "score": 7,
      "issues": {
        "content": 3,
        "terms": 3,
        "format": 2,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 363.79
    },
    "v36": {
      "n": 8,
      "passed": 0,
      "partial": 6,
      "failed": 2,
      "score": 6,
      "issues": {
        "content": 4,
        "format": 3,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 483.96
    },
    "v37": {
      "n": 8,
      "passed": 0,
      "partial": 5,
      "failed": 3,
      "score": 5,
      "issues": {
        "content": 4,
        "terms": 3,
        "numbers": 2,
        "format": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 462.57
    },
    "v38": {
      "n": 8,
      "passed": 0,
      "partial": 6,
      "failed": 2,
      "score": 6,
      "issues": {
        "content": 4,
        "format": 3,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 456.38
    },
    "v39": {
      "n": 8,
      "passed": 1,
      "partial": 4,
      "failed": 3,
      "score": 6,
      "issues": {
        "content": 3,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 376.98
    },
    "v40": {
      "n": 8,
      "passed": 0,
      "partial": 5,
      "failed": 3,
      "score": 5,
      "issues": {
        "content": 4,
        "terms": 3,
        "format": 2,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 390.16
    },
    "v41": {
      "n": 8,
      "passed": 0,
      "partial": 3,
      "failed": 5,
      "score": 3,
      "issues": {
        "format": 4,
        "content": 4,
        "numbers": 2,
        "terms": 2,
        "display": 1,
        "repetition": 1,
        "quote": 1
      },
      "median_ms": 388.22
    },
    "v42": {
      "n": 8,
      "passed": 1,
      "partial": 3,
      "failed": 4,
      "score": 5,
      "issues": {
        "repetition": 4,
        "format": 3,
        "content": 3,
        "numbers": 2,
        "display": 1,
        "terms": 1,
        "quote": 1
      },
      "median_ms": 648.66
    },
    "v43": {
      "n": 8,
      "passed": 1,
      "partial": 3,
      "failed": 4,
      "score": 5,
      "issues": {
        "content": 3,
        "format": 2,
        "repetition": 2,
        "terms": 2,
        "numbers": 2,
        "display": 1,
        "fidelity": 1
      },
      "median_ms": 526.79
    },
    "v44": {
      "n": 8,
      "passed": 0,
      "partial": 2,
      "failed": 6,
      "score": 2,
      "issues": {
        "content": 3,
        "fidelity": 2,
        "numbers": 2,
        "repetition": 2,
        "display": 1,
        "terms": 1,
        "quote": 1,
        "format": 1
      },
      "median_ms": 613.15
    },
    "v45": {
      "n": 8,
      "passed": 0,
      "partial": 6,
      "failed": 2,
      "score": 6,
      "issues": {
        "format": 5,
        "content": 4,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "quote": 1
      },
      "median_ms": 408.87
    },
    "v46": {
      "n": 8,
      "passed": 0,
      "partial": 4,
      "failed": 4,
      "score": 4,
      "issues": {
        "format": 5,
        "content": 3,
        "repetition": 3,
        "terms": 3,
        "display": 1,
        "quote": 1,
        "numbers": 1
      },
      "median_ms": 701.54
    },
    "v47": {
      "n": 8,
      "passed": 1,
      "partial": 3,
      "failed": 4,
      "score": 5,
      "issues": {
        "content": 3,
        "repetition": 2,
        "terms": 2,
        "display": 1,
        "numbers": 1,
        "fidelity": 1
      },
      "median_ms": 439.07
    },
    "v48": {
      "n": 8,
      "passed": 1,
      "partial": 3,
      "failed": 4,
      "score": 5,
      "issues": {
        "content": 3,
        "format": 2,
        "terms": 2,
        "numbers": 2,
        "display": 1,
        "fidelity": 1,
        "repetition": 1,
        "quote": 1
      },
      "median_ms": 539.63
    },
    "v49": {
      "n": 8,
      "passed": 0,
      "partial": 6,
      "failed": 2,
      "score": 6,
      "issues": {
        "content": 4,
        "format": 3,
        "terms": 3,
        "numbers": 2,
        "display": 1,
        "repetition": 1
      },
      "median_ms": 426.54
    },
    "v50": {
      "n": 8,
      "passed": 1,
      "partial": 4,
      "failed": 3,
      "score": 6,
      "issues": {
        "content": 3,
        "terms": 2,
        "numbers": 2,
        "display": 1,
        "fidelity": 1,
        "quote": 1
      },
      "median_ms": 408.79
    }
  },
  "finalists": [
    "v32",
    "v35",
    "v26",
    "v18",
    "v33",
    "v22"
  ],
  "combined": {
    "v32": {
      "n": 20,
      "passed": 3,
      "partial": 16,
      "failed": 1,
      "score": 22,
      "issues": {
        "terms": 8,
        "numbers": 5,
        "content": 4,
        "format": 2,
        "quote": 1,
        "display": 1
      },
      "median_ms": 402.15
    },
    "v35": {
      "n": 20,
      "passed": 3,
      "partial": 15,
      "failed": 2,
      "score": 21,
      "issues": {
        "terms": 7,
        "numbers": 5,
        "content": 3,
        "format": 2,
        "display": 2,
        "quote": 1,
        "fidelity": 1,
        "punctuation": 1
      },
      "median_ms": 365.23
    },
    "v26": {
      "n": 20,
      "passed": 3,
      "partial": 13,
      "failed": 4,
      "score": 19,
      "issues": {
        "terms": 8,
        "numbers": 5,
        "content": 3,
        "format": 2,
        "display": 2,
        "quote": 1,
        "punctuation": 1
      },
      "median_ms": 359.3
    },
    "v18": {
      "n": 20,
      "passed": 3,
      "partial": 14,
      "failed": 3,
      "score": 20,
      "issues": {
        "terms": 8,
        "numbers": 5,
        "content": 4,
        "display": 2,
        "format": 2,
        "quote": 1,
        "punctuation": 1
      },
      "median_ms": 365.4
    },
    "v33": {
      "n": 20,
      "passed": 4,
      "partial": 11,
      "failed": 5,
      "score": 19,
      "issues": {
        "terms": 7,
        "format": 6,
        "numbers": 5,
        "content": 3,
        "display": 2,
        "quote": 1,
        "fidelity": 1
      },
      "median_ms": 370.08
    },
    "v22": {
      "n": 20,
      "passed": 3,
      "partial": 14,
      "failed": 3,
      "score": 20,
      "issues": {
        "terms": 8,
        "numbers": 5,
        "content": 4,
        "display": 2,
        "format": 2,
        "quote": 1,
        "punctuation": 1
      },
      "median_ms": 389.05
    }
  },
  "holdout": {
    "n": 30,
    "passed": 3,
    "partial": 24,
    "failed": 3,
    "score": 30,
    "issues": {
      "numbers": 9,
      "content": 7,
      "terms": 6,
      "punctuation": 4,
      "fidelity": 3,
      "display": 3,
      "quote": 2,
      "format": 2
    },
    "median_ms": 354.24
  },
  "full": {
    "qwen35-2b": {
      "n": 50,
      "passed": 6,
      "partial": 40,
      "failed": 4,
      "score": 52,
      "issues": {
        "terms": 14,
        "numbers": 14,
        "content": 11,
        "display": 4,
        "punctuation": 4,
        "format": 4,
        "fidelity": 3,
        "quote": 3
      },
      "median_ms": 358.94
    },
    "gemma4-e2b": {
      "n": 50,
      "passed": 15,
      "partial": 18,
      "failed": 17,
      "score": 48,
      "issues": {
        "numbers": 10,
        "content": 8,
        "fidelity": 7,
        "display": 6,
        "terms": 6,
        "punctuation": 3,
        "quote": 2
      },
      "median_ms": 350.73
    }
  }
}
