{
  "benchmark_date": "2026-08-21",
  "corpus": {
    "annotated_fields": 2130,
    "documents": 23,
    "ground_truth_sha256": "c67439f5fc9013cbdbbf283b4a2e29623bd20a52af501e622fab6ad563b5d9b4"
  },
  "documents": [
    {
      "annotated_fields": 71,
      "category": "Medical intake",
      "f1": 0.0526,
      "fn": 69,
      "fp": 3,
      "id": "patient-intake-c",
      "label": "Patient intake form C",
      "precision": 0.4,
      "predictions": 5,
      "recall": 0.0282,
      "tp": 2
    },
    {
      "annotated_fields": 8,
      "category": "Government and tax",
      "f1": 0.125,
      "fn": 6,
      "fp": 22,
      "id": "new-zealand-ir330",
      "label": "New Zealand IR330",
      "precision": 0.0833,
      "predictions": 24,
      "recall": 0.25,
      "tp": 2
    },
    {
      "annotated_fields": 13,
      "category": "Banking",
      "f1": 0.1429,
      "fn": 12,
      "fp": 0,
      "id": "bank-direct-deposit-b",
      "label": "Bank direct-deposit form B",
      "precision": 1.0,
      "predictions": 1,
      "recall": 0.0769,
      "tp": 1
    },
    {
      "annotated_fields": 138,
      "category": "Government and tax",
      "f1": 0.3886,
      "fn": 104,
      "fp": 3,
      "id": "france-cerfa-13750",
      "label": "France Cerfa 13750",
      "precision": 0.9189,
      "predictions": 37,
      "recall": 0.2464,
      "tp": 34
    },
    {
      "annotated_fields": 23,
      "category": "Government and tax",
      "f1": 0.4364,
      "fn": 11,
      "fp": 20,
      "id": "irs-w-9",
      "label": "IRS W-9",
      "precision": 0.375,
      "predictions": 32,
      "recall": 0.5217,
      "tp": 12
    },
    {
      "annotated_fields": 45,
      "category": "Government and tax",
      "f1": 0.5088,
      "fn": 16,
      "fp": 40,
      "id": "hmrc-starter-checklist",
      "label": "HMRC Starter Checklist",
      "precision": 0.4203,
      "predictions": 69,
      "recall": 0.6444,
      "tp": 29
    },
    {
      "annotated_fields": 39,
      "category": "Banking",
      "f1": 0.6197,
      "fn": 17,
      "fp": 10,
      "id": "bank-direct-deposit-a",
      "label": "Bank direct-deposit form A",
      "precision": 0.6875,
      "predictions": 32,
      "recall": 0.5641,
      "tp": 22
    },
    {
      "annotated_fields": 30,
      "category": "Business form",
      "f1": 0.7297,
      "fn": 3,
      "fp": 17,
      "id": "deposit-receipt",
      "label": "Deposit receipt",
      "precision": 0.6136,
      "predictions": 44,
      "recall": 0.9,
      "tp": 27
    },
    {
      "annotated_fields": 31,
      "category": "Government and tax",
      "f1": 0.7667,
      "fn": 8,
      "fp": 6,
      "id": "canada-td1",
      "label": "Canada TD1",
      "precision": 0.7931,
      "predictions": 29,
      "recall": 0.7419,
      "tp": 23
    },
    {
      "annotated_fields": 63,
      "category": "Medical intake",
      "f1": 0.781,
      "fn": 22,
      "fp": 1,
      "id": "patient-intake-d",
      "label": "Patient intake form D",
      "precision": 0.9762,
      "predictions": 42,
      "recall": 0.6508,
      "tp": 41
    },
    {
      "annotated_fields": 104,
      "category": "Business form",
      "f1": 0.809,
      "fn": 32,
      "fp": 2,
      "id": "bid-proposal",
      "label": "Bid proposal",
      "precision": 0.973,
      "predictions": 74,
      "recall": 0.6923,
      "tp": 72
    },
    {
      "annotated_fields": 229,
      "category": "Real estate",
      "f1": 0.8235,
      "fn": 54,
      "fp": 21,
      "id": "residential-lease",
      "label": "Residential lease",
      "precision": 0.8929,
      "predictions": 196,
      "recall": 0.7642,
      "tp": 175
    },
    {
      "annotated_fields": 48,
      "category": "Government and tax",
      "f1": 0.8261,
      "fn": 10,
      "fp": 6,
      "id": "irs-w-4",
      "label": "IRS W-4",
      "precision": 0.8636,
      "predictions": 44,
      "recall": 0.7917,
      "tp": 38
    },
    {
      "annotated_fields": 125,
      "category": "Business form",
      "f1": 0.8547,
      "fn": 25,
      "fp": 9,
      "id": "contractor-invoice",
      "label": "Contractor invoice",
      "precision": 0.9174,
      "predictions": 109,
      "recall": 0.8,
      "tp": 100
    },
    {
      "annotated_fields": 41,
      "category": "Agreement",
      "f1": 0.8706,
      "fn": 4,
      "fp": 7,
      "id": "power-of-attorney",
      "label": "Power of attorney",
      "precision": 0.8409,
      "predictions": 44,
      "recall": 0.9024,
      "tp": 37
    },
    {
      "annotated_fields": 83,
      "category": "Agreement",
      "f1": 0.8718,
      "fn": 15,
      "fp": 5,
      "id": "car-rental-agreement",
      "label": "Car rental agreement",
      "precision": 0.9315,
      "predictions": 73,
      "recall": 0.8193,
      "tp": 68
    },
    {
      "annotated_fields": 31,
      "category": "Government and tax",
      "f1": 0.8727,
      "fn": 7,
      "fp": 0,
      "id": "us-court-ao-240",
      "label": "U.S. Court AO 240",
      "precision": 1.0,
      "predictions": 24,
      "recall": 0.7742,
      "tp": 24
    },
    {
      "annotated_fields": 176,
      "category": "Medical intake",
      "f1": 0.8847,
      "fn": 34,
      "fp": 3,
      "id": "patient-intake-b",
      "label": "Patient intake form B",
      "precision": 0.9793,
      "predictions": 145,
      "recall": 0.8068,
      "tp": 142
    },
    {
      "annotated_fields": 112,
      "category": "Agreement",
      "f1": 0.9041,
      "fn": 13,
      "fp": 8,
      "id": "investor-agreement",
      "label": "Investor agreement",
      "precision": 0.9252,
      "predictions": 107,
      "recall": 0.8839,
      "tp": 99
    },
    {
      "annotated_fields": 135,
      "category": "Agreement",
      "f1": 0.9225,
      "fn": 16,
      "fp": 4,
      "id": "independent-contractor-agreement",
      "label": "Independent contractor agreement",
      "precision": 0.9675,
      "predictions": 123,
      "recall": 0.8815,
      "tp": 119
    },
    {
      "annotated_fields": 120,
      "category": "Agreement",
      "f1": 0.9746,
      "fn": 5,
      "fp": 1,
      "id": "work-for-hire-agreement",
      "label": "Work-for-hire agreement",
      "precision": 0.9914,
      "predictions": 116,
      "recall": 0.9583,
      "tp": 115
    },
    {
      "annotated_fields": 109,
      "category": "Agreement",
      "f1": 0.9815,
      "fn": 3,
      "fp": 1,
      "id": "collaboration-agreement",
      "label": "Collaboration agreement",
      "precision": 0.9907,
      "predictions": 107,
      "recall": 0.9725,
      "tp": 106
    },
    {
      "annotated_fields": 356,
      "category": "Medical intake",
      "f1": 0.9944,
      "fn": 0,
      "fp": 4,
      "id": "patient-intake-a",
      "label": "Patient intake form A",
      "precision": 0.9889,
      "predictions": 360,
      "recall": 1.0,
      "tp": 356
    }
  ],
  "method": {
    "evaluator_schema": 4,
    "iou_threshold": 0.5,
    "macro_definition": "unweighted mean of per-document F1 recomputed from raw TP, FP, and FN",
    "matching": "one-to-one rectangle matching on the same page",
    "metric": "geometry-only precision, recall, and harmonic-mean F1",
    "scope": "flat-digital fallback detector; field type ignored"
  },
  "overall": {
    "annotated_fields": 2130,
    "documents": 23,
    "f1": 0.8288,
    "fn": 486,
    "fp": 193,
    "macro_f1": 0.7018,
    "median_f1": 0.8235,
    "precision": 0.8949,
    "predictions": 1837,
    "recall": 0.7718,
    "tp": 1644
  },
  "subsets": {
    "expansion_15": {
      "annotated_fields": 1396,
      "documents": 15,
      "f1": 0.7866,
      "fn": 395,
      "fp": 148,
      "macro_f1": 0.6052,
      "median_f1": 0.7667,
      "precision": 0.8712,
      "predictions": 1149,
      "recall": 0.717,
      "tp": 1001
    },
    "government_tax_7": {
      "annotated_fields": 324,
      "documents": 7,
      "f1": 0.5557,
      "fn": 162,
      "fp": 97,
      "macro_f1": 0.5606,
      "median_f1": 0.5088,
      "precision": 0.6255,
      "predictions": 259,
      "recall": 0.5,
      "tp": 162
    },
    "pre_expansion_8": {
      "annotated_fields": 734,
      "documents": 8,
      "f1": 0.9044,
      "fn": 91,
      "fp": 45,
      "macro_f1": 0.883,
      "median_f1": 0.888,
      "precision": 0.9346,
      "predictions": 688,
      "recall": 0.876,
      "tp": 643
    },
    "validation_3": {
      "annotated_fields": 260,
      "documents": 3,
      "f1": 0.7628,
      "fn": 67,
      "fp": 53,
      "macro_f1": 0.6711,
      "median_f1": 0.6197,
      "precision": 0.7846,
      "predictions": 246,
      "recall": 0.7423,
      "tp": 193
    },
    "without_targeted_patient": {
      "annotated_fields": 1774,
      "documents": 22,
      "f1": 0.7924,
      "fn": 486,
      "fp": 189,
      "macro_f1": 0.6885,
      "median_f1": 0.8163,
      "precision": 0.872,
      "predictions": 1477,
      "recall": 0.726,
      "tp": 1288
    }
  }
}