{
 "typesafe/jev-1.13|intent8": {
  "n": 160,
  "answered": 160,
  "failed": 0,
  "acc_of_all": 0.8375,
  "acc_ci": [
   0.7725461426123724,
   0.8866271076488556
  ],
  "lat_p50": 0.3271822914994118,
  "lat_p95": 0.7037164579996897,
  "lat_mean": 0.36088495024376926,
  "cost_per_1k": 0.015114749999999986,
  "in_tok_mean": 359.875,
  "gating": [
   {
    "min_conf": 0.0,
    "coverage": 1.0,
    "accuracy": 0.8375
   },
   {
    "min_conf": 0.5,
    "coverage": 0.9625,
    "accuracy": 0.8636363636363636
   },
   {
    "min_conf": 0.7,
    "coverage": 0.875,
    "accuracy": 0.9071428571428571
   },
   {
    "min_conf": 0.8,
    "coverage": 0.80625,
    "accuracy": 0.937984496124031
   },
   {
    "min_conf": 0.9,
    "coverage": 0.7,
    "accuracy": 0.9553571428571429
   },
   {
    "min_conf": 0.95,
    "coverage": 0.6,
    "accuracy": 0.96875
   }
  ]
 },
 "typesafe/jev-1.13|intent77": {
  "n": 231,
  "answered": 231,
  "failed": 0,
  "acc_of_all": 0.7878787878787878,
  "acc_ci": [
   0.7306713640867136,
   0.8356678194310277
  ],
  "lat_p50": 0.322906833000161,
  "lat_p95": 0.4063934170007997,
  "lat_mean": 0.34221305139395886,
  "cost_per_1k": 0.03996890909090907,
  "in_tok_mean": 951.6406926406927,
  "gating": [
   {
    "min_conf": 0.0,
    "coverage": 1.0,
    "accuracy": 0.7878787878787878
   },
   {
    "min_conf": 0.5,
    "coverage": 0.922077922077922,
    "accuracy": 0.8028169014084507
   },
   {
    "min_conf": 0.7,
    "coverage": 0.8268398268398268,
    "accuracy": 0.8534031413612565
   },
   {
    "min_conf": 0.8,
    "coverage": 0.7705627705627706,
    "accuracy": 0.8820224719101124
   },
   {
    "min_conf": 0.9,
    "coverage": 0.6623376623376623,
    "accuracy": 0.9215686274509803
   },
   {
    "min_conf": 0.95,
    "coverage": 0.5670995670995671,
    "accuracy": 0.9541984732824428
   }
  ]
 },
 "typesafe/jev-1.13|inject": {
  "n": 400,
  "answered": 400,
  "failed": 0,
  "acc_of_all": 0.87,
  "acc_ci": [
   0.8334914191139058,
   0.8994692268423305
  ],
  "lat_p50": 0.3269763750004131,
  "lat_p95": 0.420287749999261,
  "lat_mean": 0.34260778246752577,
  "cost_per_1k": 0.013603064999999977,
  "in_tok_mean": 323.8825,
  "tp": 118,
  "fp": 0,
  "fn": 52,
  "tn": 230,
  "precision": 1.0,
  "recall": 0.6941176470588235,
  "f1": 0.8194444444444444,
  "auroc": 0.9902557544757034,
  "brier": 0.08951525
 },
 "openai/gpt-5.4-nano|intent8": {
  "n": 160,
  "answered": 160,
  "failed": 0,
  "acc_of_all": 0.9,
  "acc_ci": [
   0.843736268117816,
   0.9375061025621578
  ],
  "lat_p50": 1.2188566459999493,
  "lat_p95": 2.0832807080005296,
  "lat_mean": 1.3975913632124617,
  "cost_per_1k": 0.07043014687500004,
  "in_tok_mean": 152.54375
 },
 "openai/gpt-5.4-nano|intent77": {
  "n": 231,
  "answered": 231,
  "failed": 0,
  "acc_of_all": 0.7835497835497836,
  "acc_ci": [
   0.726035437522837,
   0.8317873672953894
  ],
  "lat_p50": 1.0732669579992944,
  "lat_p95": 2.1963884580000013,
  "lat_mean": 1.2910721561948655,
  "cost_per_1k": 0.2049392142857141,
  "in_tok_mean": 828.3636363636364
 },
 "openai/gpt-5.4-nano|inject": {
  "n": 400,
  "answered": 400,
  "failed": 0,
  "acc_of_all": 0.8075,
  "acc_ci": [
   0.766010151768214,
   0.8431395742629825
  ],
  "lat_p50": 1.1561496459999034,
  "lat_p95": 1.901183874999333,
  "lat_mean": 1.2666822284675208,
  "cost_per_1k": 0.08060827500000003,
  "in_tok_mean": 126.925,
  "tp": 93,
  "fp": 0,
  "fn": 77,
  "tn": 230,
  "precision": 1.0,
  "recall": 0.5470588235294118,
  "f1": 0.7072243346007605,
  "auroc": 0.9496803069053709,
  "brier": 0.17283175
 },
 "google/gemini-3.5-flash-lite|intent8": {
  "n": 160,
  "answered": 160,
  "failed": 0,
  "acc_of_all": 0.775,
  "acc_ci": [
   0.7042857994110268,
   0.8328183304314551
  ],
  "lat_p50": 0.6669900205006343,
  "lat_p95": 0.9187125000007654,
  "lat_mean": 0.7423286473876146,
  "cost_per_1k": 0.08693994375,
  "in_tok_mean": 186.425
 },
 "google/gemini-3.5-flash-lite|intent77": {
  "n": 231,
  "answered": 231,
  "failed": 0,
  "acc_of_all": 0.7835497835497836,
  "acc_ci": [
   0.726035437522837,
   0.8317873672953894
  ],
  "lat_p50": 0.6684105830008775,
  "lat_p95": 0.9475757500003965,
  "lat_mean": 0.7599949233419812,
  "cost_per_1k": 0.30087171428571413,
  "in_tok_mean": 928.8008658008658
 },
 "google/gemini-3.5-flash-lite|inject": {
  "n": 400,
  "answered": 400,
  "failed": 0,
  "acc_of_all": 0.8675,
  "acc_ci": [
   0.830752941175613,
   0.8972552679836218
  ],
  "lat_p50": 0.6641903545005334,
  "lat_p95": 0.8346697920005681,
  "lat_mean": 0.7113099362375033,
  "cost_per_1k": 0.08600649750000007,
  "in_tok_mean": 150.0425,
  "tp": 117,
  "fp": 0,
  "fn": 53,
  "tn": 230,
  "precision": 1.0,
  "recall": 0.6882352941176471,
  "f1": 0.8153310104529616,
  "auroc": 0.8470588235294118,
  "brier": 0.13275575
 },
 "anthropic/claude-haiku-4.5|intent8": {
  "n": 160,
  "answered": 160,
  "failed": 0,
  "acc_of_all": 0.7875,
  "acc_ci": [
   0.7177577354595728,
   0.8437602184666584
  ],
  "lat_p50": 0.9917887080000583,
  "lat_p95": 1.6863758750005218,
  "lat_mean": 1.0935410557313616,
  "cost_per_1k": 0.35646806249999996,
  "in_tok_mean": 303.00625
 },
 "anthropic/claude-haiku-4.5|intent77": {
  "n": 231,
  "answered": 231,
  "failed": 0,
  "acc_of_all": 0.7619047619047619,
  "acc_ci": [
   0.7029776141838151,
   0.8122632971368366
  ],
  "lat_p50": 1.0107703749999928,
  "lat_p95": 1.9794130000009318,
  "lat_mean": 1.119956357142868,
  "cost_per_1k": 1.2550499999999998,
  "in_tok_mean": 1210.9307359307359
 },
 "anthropic/claude-haiku-4.5|inject": {
  "n": 400,
  "answered": 400,
  "failed": 0,
  "acc_of_all": 0.89,
  "acc_ci": [
   0.8555483868341889,
   0.9170317534980603
  ],
  "lat_p50": 1.0329209584997443,
  "lat_p95": 1.947362292001344,
  "lat_mean": 1.136943948234948,
  "cost_per_1k": 0.3438517500000002,
  "in_tok_mean": 267.325,
  "tp": 126,
  "fp": 0,
  "fn": 44,
  "tn": 230,
  "precision": 1.0,
  "recall": 0.7411764705882353,
  "f1": 0.8513513513513514,
  "auroc": 0.9815601023017902,
  "brier": 0.09826625
 },
 "openai/gpt-5.6-terra|intent8": {
  "n": 160,
  "answered": 160,
  "failed": 0,
  "acc_of_all": 0.89375,
  "acc_ci": [
   0.8364366789581195,
   0.9325987796799797
  ],
  "lat_p50": 1.1840447499998845,
  "lat_p95": 2.721009540999148,
  "lat_mean": 1.577978448437534,
  "cost_per_1k": 0.6089118750000001,
  "in_tok_mean": 152.54375
 },
 "openai/gpt-5.6-terra|intent77": {
  "n": 231,
  "answered": 230,
  "failed": 1,
  "acc_of_all": 0.8398268398268398,
  "acc_ci": [
   0.7870303993609443,
   0.8815053285509759
  ],
  "lat_p50": 1.1048419785001897,
  "lat_p95": 2.94273025000075,
  "lat_mean": 1.3929360335086725,
  "cost_per_1k": 1.9568856521739126,
  "in_tok_mean": 828.3869565217391
 },
 "openai/gpt-5.6-terra|inject": {
  "n": 400,
  "answered": 400,
  "failed": 0,
  "acc_of_all": 0.8675,
  "acc_ci": [
   0.830752941175613,
   0.8972552679836218
  ],
  "lat_p50": 1.2223743130007279,
  "lat_p95": 2.3278401249990566,
  "lat_mean": 1.3659632647950366,
  "cost_per_1k": 0.6733138500000002,
  "in_tok_mean": 126.925,
  "tp": 117,
  "fp": 0,
  "fn": 53,
  "tn": 230,
  "precision": 1.0,
  "recall": 0.6882352941176471,
  "f1": 0.8153310104529616,
  "auroc": 0.9249616368286445,
  "brier": 0.107084
 }
}