EmojiFinder / evaluation.json
Vicente1313's picture
Release EmojiFinder
69ee812
Raw History Blame Contribute Delete
9 kB
{
"recorded_date": "2026-10-02",
"checkpoint_sha256": "642eb7a64bfc95be17c8ed536e71a9d695eee79a13449847563f28644cf7e7ac",
"baseline": "development reference",
"candidate": "EmojiFinder",
"sets": {
"user": {
"before": {
"precision": 1.0,
"recall": 1.0,
"f05": 1.0,
"extra": 0,
"missing": 0,
"queries": 20,
"exact": 20,
"queries_with_extras": 0,
"first_ok": 20,
"first_total": 20,
"primary_order_ok": 20
},
"after": {
"precision": 1.0,
"recall": 1.0,
"f05": 1.0,
"extra": 0,
"missing": 0,
"queries": 20,
"exact": 20,
"queries_with_extras": 0,
"first_ok": 20,
"first_total": 20,
"primary_order_ok": 20
},
"queries": 20
},
"probes": {
"before": {
"precision": 1.0,
"recall": 1.0,
"f05": 1.0,
"extra": 0,
"missing": 0,
"queries": 191,
"exact": 191,
"queries_with_extras": 0,
"first_ok": 36,
"first_total": 36,
"primary_order_ok": 36
},
"after": {
"precision": 1.0,
"recall": 1.0,
"f05": 1.0,
"extra": 0,
"missing": 0,
"queries": 191,
"exact": 191,
"queries_with_extras": 0,
"first_ok": 36,
"first_total": 36,
"primary_order_ok": 36
},
"queries": 191
},
"reviewed": {
"before": {
"precision": 1.0,
"recall": 1.0,
"f05": 1.0,
"extra": 0,
"missing": 0,
"queries": 1839,
"exact": 1839,
"queries_with_extras": 0,
"first_ok": 1660,
"first_total": 1660,
"primary_order_ok": 1660
},
"after": {
"precision": 1.0,
"recall": 1.0,
"f05": 1.0,
"extra": 0,
"missing": 0,
"queries": 1839,
"exact": 1839,
"queries_with_extras": 0,
"first_ok": 1660,
"first_total": 1660,
"primary_order_ok": 1660
},
"queries": 1839
},
"validation": {
"before": {
"precision": 0.6142857142857143,
"recall": 0.7478260869565218,
"f05": 0.6370370370370371,
"extra": 54,
"missing": 29,
"queries": 111,
"exact": 68,
"queries_with_extras": 23,
"first_ok": 84,
"first_total": 111,
"primary_order_ok": 111
},
"after": {
"precision": 0.6027397260273972,
"recall": 0.7652173913043478,
"f05": 0.6294706723891274,
"extra": 58,
"missing": 27,
"queries": 111,
"exact": 67,
"queries_with_extras": 24,
"first_ok": 84,
"first_total": 111,
"primary_order_ok": 111
},
"queries": 111
},
"test": {
"before": {
"precision": 0.5562913907284768,
"recall": 0.717948717948718,
"f05": 0.5825242718446603,
"extra": 67,
"missing": 33,
"queries": 112,
"exact": 61,
"queries_with_extras": 24,
"first_ok": 74,
"first_total": 112,
"primary_order_ok": 112
},
"after": {
"precision": 0.5838926174496645,
"recall": 0.7435897435897436,
"f05": 0.6100981767180926,
"extra": 62,
"missing": 30,
"queries": 112,
"exact": 64,
"queries_with_extras": 22,
"first_ok": 81,
"first_total": 112,
"primary_order_ok": 112
},
"queries": 112
},
"old_validation": {
"before": {
"precision": 0.9981782605762094,
"recall": 0.9908244591788895,
"f05": 0.9966987805699656,
"extra": 27,
"missing": 137,
"queries": 3037,
"exact": 2965,
"queries_with_extras": 20,
"first_ok": 0,
"first_total": 0,
"primary_order_ok": 0
},
"after": {
"precision": 0.9988487844518182,
"recall": 0.9878775701560512,
"f05": 0.9966350896633738,
"extra": 17,
"missing": 181,
"queries": 3037,
"exact": 2962,
"queries_with_extras": 16,
"first_ok": 0,
"first_total": 0,
"primary_order_ok": 0
},
"queries": 3037
},
"old_test": {
"before": {
"precision": 0.9986547386829892,
"recall": 0.99530736743313,
"f05": 0.9979834644081467,
"extra": 20,
"missing": 70,
"queries": 3111,
"exact": 3059,
"queries_with_extras": 16,
"first_ok": 0,
"first_total": 0,
"primary_order_ok": 0
},
"after": {
"precision": 0.9982482145263442,
"recall": 0.9932292015820876,
"f05": 0.9972403580803662,
"extra": 26,
"missing": 101,
"queries": 3111,
"exact": 3052,
"queries_with_extras": 19,
"first_ok": 0,
"first_total": 0,
"primary_order_ok": 0
},
"queries": 3111
},
"previous_test": {
"before": {
"precision": 0.9985980804486143,
"recall": 0.9966634377354429,
"f05": 0.9982105512795636,
"extra": 13,
"missing": 31,
"queries": 2944,
"exact": 2923,
"queries_with_extras": 11,
"first_ok": 0,
"first_total": 0,
"primary_order_ok": 0
},
"after": {
"precision": 0.9991384880465216,
"recall": 0.9986007964697019,
"f05": 0.9990309034133734,
"extra": 8,
"missing": 13,
"queries": 2944,
"exact": 2930,
"queries_with_extras": 8,
"first_ok": 0,
"first_total": 0,
"primary_order_ok": 0
},
"queries": 2944
},
"previous_phrase_validation": {
"before": {
"precision": 0.9983265598852499,
"recall": 0.9870007090522335,
"f05": 0.9960406430377332,
"extra": 7,
"missing": 55,
"queries": 1629,
"exact": 1606,
"queries_with_extras": 6,
"first_ok": 1447,
"first_total": 1452,
"primary_order_ok": 1445
},
"after": {
"precision": 0.9820431933996603,
"recall": 0.9565114630111085,
"f05": 0.9768283852280956,
"extra": 74,
"missing": 184,
"queries": 1629,
"exact": 1533,
"queries_with_extras": 23,
"first_ok": 1414,
"first_total": 1452,
"primary_order_ok": 1422
},
"queries": 1629
},
"previous_phrase_test": {
"before": {
"precision": 0.9966547192353644,
"recall": 0.9858189553297093,
"f05": 0.994468551809642,
"extra": 14,
"missing": 60,
"queries": 1629,
"exact": 1592,
"queries_with_extras": 6,
"first_ok": 1446,
"first_total": 1452,
"primary_order_ok": 1444
},
"after": {
"precision": 0.9615384615384616,
"recall": 0.9926731269203498,
"f05": 0.9676081647698475,
"extra": 168,
"missing": 31,
"queries": 1629,
"exact": 1600,
"queries_with_extras": 22,
"first_ok": 1441,
"first_total": 1452,
"primary_order_ok": 1445
},
"queries": 1629
}
},
"categories": {
"train": {
"before": {
"queries": 148,
"passed": 9,
"primary_first": 85,
"primary_before_related": 22,
"required_total": 2466,
"missing_total": 2063,
"forbidden_total": 0
},
"after": {
"queries": 148,
"passed": 148,
"primary_first": 148,
"primary_before_related": 148,
"required_total": 2466,
"missing_total": 0,
"forbidden_total": 0
}
},
"validation": {
"before": {
"queries": 64,
"passed": 0,
"primary_first": 27,
"primary_before_related": 4,
"required_total": 1048,
"missing_total": 926,
"forbidden_total": 0
},
"after": {
"queries": 64,
"passed": 11,
"primary_first": 46,
"primary_before_related": 19,
"required_total": 1048,
"missing_total": 512,
"forbidden_total": 0
}
},
"test": {
"before": {
"queries": 64,
"passed": 0,
"primary_first": 24,
"primary_before_related": 5,
"required_total": 1048,
"missing_total": 943,
"forbidden_total": 0
},
"after": {
"queries": 64,
"passed": 0,
"primary_first": 32,
"primary_before_related": 7,
"required_total": 1048,
"missing_total": 842,
"forbidden_total": 0
}
}
},
"limitations": "Assistant-authored labels, overlapping regression suites and prior test exposure; not blind human evaluation."
}