{
  "train_key": "456965ab480be4a4",
  "finetuned": {
    "all": {
      "n": 89838,
      "accuracy": 0.6809,
      "brier": 0.4125,
      "nll": 0.8885,
      "ece": 0.0701
    },
    "source/bev_tr": {
      "n": 46320,
      "accuracy": 0.771,
      "brier": 0.3141,
      "nll": 0.589,
      "ece": 0.0314
    },
    "source/bev_tr/choice": {
      "n": 18808,
      "accuracy": 0.7402,
      "brier": 0.3458,
      "nll": 0.6765,
      "ece": 0.0545
    },
    "source/bev_tr/noul": {
      "n": 17572,
      "accuracy": 0.8692,
      "brier": 0.2012,
      "nll": 0.3362,
      "ece": 0.0296
    },
    "source/bev_tr/score": {
      "n": 9940,
      "accuracy": 0.6557,
      "brier": 0.4537,
      "nll": 0.87,
      "ece": 0.0524
    },
    "source/bev_tr_en": {
      "n": 5761,
      "accuracy": 0.775,
      "brier": 0.3109,
      "nll": 0.5888,
      "ece": 0.0363
    },
    "source/bev_tr_en/choice": {
      "n": 2387,
      "accuracy": 0.7357,
      "brier": 0.3578,
      "nll": 0.7025,
      "ece": 0.0551
    },
    "source/bev_tr_en/noul": {
      "n": 2112,
      "accuracy": 0.8845,
      "brier": 0.1782,
      "nll": 0.3048,
      "ece": 0.027
    },
    "source/bev_tr_en/score": {
      "n": 1262,
      "accuracy": 0.6664,
      "brier": 0.4441,
      "nll": 0.849,
      "ece": 0.0586
    },
    "source/beyazperde_tr": {
      "n": 5000,
      "accuracy": 0.5312,
      "brier": 0.5973,
      "nll": 1.1068,
      "ece": 0.0665
    },
    "source/beyazperde_tr/score": {
      "n": 5000,
      "accuracy": 0.5312,
      "brier": 0.5973,
      "nll": 1.1068,
      "ece": 0.0665
    },
    "source/jev_bench_tr": {
      "n": 22773,
      "accuracy": 0.4902,
      "brier": 0.6055,
      "nll": 1.5258,
      "ece": 0.1707
    },
    "source/jev_bench_tr/choice": {
      "n": 9599,
      "accuracy": 0.4262,
      "brier": 0.6506,
      "nll": 2.1598,
      "ece": 0.1854
    },
    "source/jev_bench_tr/noul": {
      "n": 7174,
      "accuracy": 0.6968,
      "brier": 0.4359,
      "nll": 0.7049,
      "ece": 0.1706
    },
    "source/jev_bench_tr/score": {
      "n": 6000,
      "accuracy": 0.3457,
      "brier": 0.736,
      "nll": 1.4929,
      "ece": 0.1472
    },
    "source/jev_bench_tr_en": {
      "n": 2000,
      "accuracy": 0.5775,
      "brier": 0.4968,
      "nll": 1.2451,
      "ece": 0.1054
    },
    "source/jev_bench_tr_en/choice": {
      "n": 839,
      "accuracy": 0.5459,
      "brier": 0.5285,
      "nll": 1.7378,
      "ece": 0.1067
    },
    "source/jev_bench_tr_en/noul": {
      "n": 635,
      "accuracy": 0.7795,
      "brier": 0.2983,
      "nll": 0.5095,
      "ece": 0.0858
    },
    "source/jev_bench_tr_en/score": {
      "n": 526,
      "accuracy": 0.384,
      "brier": 0.6857,
      "nll": 1.3473,
      "ece": 0.1316
    },
    "source/massive_tr": {
      "n": 2974,
      "accuracy": 0.7811,
      "brier": 0.3297,
      "nll": 0.9699,
      "ece": 0.0632
    },
    "source/massive_tr/choice": {
      "n": 2974,
      "accuracy": 0.7811,
      "brier": 0.3297,
      "nll": 0.9699,
      "ece": 0.0632
    },
    "source/xnli_tr": {
      "n": 5010,
      "accuracy": 0.7375,
      "brier": 0.3926,
      "nll": 0.6972,
      "ece": 0.0949
    },
    "source/xnli_tr/choice": {
      "n": 5010,
      "accuracy": 0.7375,
      "brier": 0.3926,
      "nll": 0.6972,
      "ece": 0.0949
    },
    "type/choice": {
      "n": 39617,
      "accuracy": 0.6625,
      "brier": 0.429,
      "nll": 1.0846,
      "ece": 0.0905
    },
    "type/noul": {
      "n": 27493,
      "accuracy": 0.8233,
      "brier": 0.2629,
      "nll": 0.434,
      "ece": 0.061
    },
    "type/score": {
      "n": 22728,
      "accuracy": 0.5408,
      "brier": 0.5647,
      "nll": 1.0964,
      "ece": 0.0713
    },
    "workflow/bev_tr/Browser interaction": {
      "n": 286,
      "accuracy": 0.7448,
      "brier": 0.3899,
      "nll": 0.7802,
      "ece": 0.0895
    },
    "workflow/bev_tr/Civic and safety operations": {
      "n": 261,
      "accuracy": 0.7011,
      "brier": 0.37,
      "nll": 0.7452,
      "ece": 0.0761
    },
    "workflow/bev_tr/Contract evidence": {
      "n": 320,
      "accuracy": 0.7,
      "brier": 0.3975,
      "nll": 0.6832,
      "ece": 0.043
    },
    "workflow/bev_tr/Engagement and ranking": {
      "n": 82,
      "accuracy": 0.5732,
      "brier": 0.4914,
      "nll": 0.697,
      "ece": 0.1592
    },
    "workflow/bev_tr/Financial reporting and banking": {
      "n": 1740,
      "accuracy": 0.5994,
      "brier": 0.5204,
      "nll": 0.9951,
      "ece": 0.0424
    },
    "workflow/bev_tr/Game-state decisions": {
      "n": 2039,
      "accuracy": 0.8558,
      "brier": 0.2074,
      "nll": 0.3641,
      "ece": 0.0701
    },
    "workflow/bev_tr/Reading comprehension": {
      "n": 1422,
      "accuracy": 0.7053,
      "brier": 0.4404,
      "nll": 0.6624,
      "ece": 0.1453
    },
    "workflow/bev_tr/Response preference and quality": {
      "n": 3176,
      "accuracy": 0.5699,
      "brier": 0.5603,
      "nll": 1.0115,
      "ece": 0.0386
    },
    "workflow/bev_tr/Retail, product, and shopping": {
      "n": 6601,
      "accuracy": 0.9382,
      "brier": 0.1036,
      "nll": 0.2381,
      "ece": 0.0639
    },
    "workflow/bev_tr/Scientific and paper understanding": {
      "n": 508,
      "accuracy": 0.8268,
      "brier": 0.2649,
      "nll": 0.4787,
      "ece": 0.0623
    },
    "workflow/bev_tr/Sensory quality rating": {
      "n": 260,
      "accuracy": 0.45,
      "brier": 0.6741,
      "nll": 1.2858,
      "ece": 0.072
    },
    "workflow/bev_tr/Sentiment, emotion, and moderation": {
      "n": 11526,
      "accuracy": 0.8525,
      "brier": 0.2205,
      "nll": 0.4172,
      "ece": 0.032
    },
    "workflow/bev_tr/Software engineering and code": {
      "n": 2039,
      "accuracy": 0.6106,
      "brier": 0.5455,
      "nll": 1.0456,
      "ece": 0.1368
    },
    "workflow/bev_tr/Software security": {
      "n": 2240,
      "accuracy": 0.6531,
      "brier": 0.4622,
      "nll": 0.792,
      "ece": 0.0884
    },
    "workflow/bev_tr/Spam detection": {
      "n": 271,
      "accuracy": 0.9557,
      "brier": 0.0816,
      "nll": 0.1752,
      "ece": 0.0468
    },
    "workflow/bev_tr/Spatial and logical reasoning": {
      "n": 8384,
      "accuracy": 0.5816,
      "brier": 0.5369,
      "nll": 0.9921,
      "ece": 0.0868
    },
    "workflow/bev_tr/Support and intent routing": {
      "n": 2858,
      "accuracy": 0.9699,
      "brier": 0.0556,
      "nll": 0.1613,
      "ece": 0.0541
    },
    "workflow/bev_tr/Tool and workflow decisions": {
      "n": 2307,
      "accuracy": 0.9857,
      "brier": 0.0322,
      "nll": 0.1165,
      "ece": 0.0651
    },
    "workflow/bev_tr_en/Browser interaction": {
      "n": 39,
      "accuracy": 0.7179,
      "brier": 0.4923,
      "nll": 0.9626,
      "ece": 0.2261
    },
    "workflow/bev_tr_en/Civic and safety operations": {
      "n": 31,
      "accuracy": 0.4839,
      "brier": 0.6397,
      "nll": 1.2239,
      "ece": 0.2769
    },
    "workflow/bev_tr_en/Contract evidence": {
      "n": 49,
      "accuracy": 0.7551,
      "brier": 0.3823,
      "nll": 0.7029,
      "ece": 0.2028
    },
    "workflow/bev_tr_en/Engagement and ranking": {
      "n": 6,
      "accuracy": 0.8333,
      "brier": 0.4349,
      "nll": 0.6269,
      "ece": 0.3409
    },
    "workflow/bev_tr_en/Financial reporting and banking": {
      "n": 196,
      "accuracy": 0.6735,
      "brier": 0.466,
      "nll": 0.94,
      "ece": 0.0875
    },
    "workflow/bev_tr_en/Game-state decisions": {
      "n": 303,
      "accuracy": 0.8977,
      "brier": 0.1682,
      "nll": 0.3118,
      "ece": 0.0653
    },
    "workflow/bev_tr_en/Reading comprehension": {
      "n": 168,
      "accuracy": 0.6786,
      "brier": 0.4621,
      "nll": 0.6826,
      "ece": 0.1802
    },
    "workflow/bev_tr_en/Response preference and quality": {
      "n": 389,
      "accuracy": 0.5116,
      "brier": 0.6053,
      "nll": 1.1036,
      "ece": 0.0985
    },
    "workflow/bev_tr_en/Retail, product, and shopping": {
      "n": 902,
      "accuracy": 0.9424,
      "brier": 0.1003,
      "nll": 0.2413,
      "ece": 0.0704
    },
    "workflow/bev_tr_en/Scientific and paper understanding": {
      "n": 66,
      "accuracy": 0.8485,
      "brier": 0.2585,
      "nll": 0.5006,
      "ece": 0.0904
    },
    "workflow/bev_tr_en/Sensory quality rating": {
      "n": 37,
      "accuracy": 0.4865,
      "brier": 0.6809,
      "nll": 1.2914,
      "ece": 0.2337
    },
    "workflow/bev_tr_en/Sentiment, emotion, and moderation": {
      "n": 1433,
      "accuracy": 0.8946,
      "brier": 0.1676,
      "nll": 0.3238,
      "ece": 0.0415
    },
    "workflow/bev_tr_en/Software engineering and code": {
      "n": 260,
      "accuracy": 0.6154,
      "brier": 0.5456,
      "nll": 1.0562,
      "ece": 0.1593
    },
    "workflow/bev_tr_en/Software security": {
      "n": 234,
      "accuracy": 0.6154,
      "brier": 0.512,
      "nll": 0.8692,
      "ece": 0.1382
    },
    "workflow/bev_tr_en/Spam detection": {
      "n": 23,
      "accuracy": 1.0,
      "brier": 0.0103,
      "nll": 0.0743,
      "ece": 0.0716
    },
    "workflow/bev_tr_en/Spatial and logical reasoning": {
      "n": 1042,
      "accuracy": 0.5384,
      "brier": 0.589,
      "nll": 1.0846,
      "ece": 0.1192
    },
    "workflow/bev_tr_en/Support and intent routing": {
      "n": 310,
      "accuracy": 0.9613,
      "brier": 0.0693,
      "nll": 0.1853,
      "ece": 0.0497
    },
    "workflow/bev_tr_en/Tool and workflow decisions": {
      "n": 273,
      "accuracy": 0.9927,
      "brier": 0.0185,
      "nll": 0.0942,
      "ece": 0.0709
    },
    "workflow/beyazperde_tr": {
      "n": 5000,
      "accuracy": 0.5312,
      "brier": 0.5973,
      "nll": 1.1068,
      "ece": 0.0665
    },
    "workflow/jev_bench_tr/arc_challenge": {
      "n": 1000,
      "accuracy": 0.254,
      "brier": 0.9528,
      "nll": 1.7954,
      "ece": 0.3298
    },
    "workflow/jev_bench_tr/banking77": {
      "n": 1000,
      "accuracy": 0.323,
      "brier": 0.9502,
      "nll": 3.3116,
      "ece": 0.2924
    },
    "workflow/jev_bench_tr/boolq": {
      "n": 1000,
      "accuracy": 0.633,
      "brier": 0.5397,
      "nll": 0.79,
      "ece": 0.2194
    },
    "workflow/jev_bench_tr/chaosnli": {
      "n": 1599,
      "accuracy": 0.5697,
      "brier": 0.3207,
      "nll": 1.3576,
      "ece": 0.2328
    },
    "workflow/jev_bench_tr/civil_comments": {
      "n": 2000,
      "accuracy": 0.8305,
      "brier": 0.1849,
      "nll": 0.4636,
      "ece": 0.0813
    },
    "workflow/jev_bench_tr/clinc150": {
      "n": 1000,
      "accuracy": 0.387,
      "brier": 0.7844,
      "nll": 2.8726,
      "ece": 0.0485
    },
    "workflow/jev_bench_tr/fever_evidence": {
      "n": 1000,
      "accuracy": 0.752,
      "brier": 0.3458,
      "nll": 0.5243,
      "ece": 0.0731
    },
    "workflow/jev_bench_tr/go_emotions": {
      "n": 1000,
      "accuracy": 0.308,
      "brier": 0.4802,
      "nll": 2.9553,
      "ece": 0.2432
    },
    "workflow/jev_bench_tr/helpsteer2_helpfulness": {
      "n": 1000,
      "accuracy": 0.156,
      "brier": 0.881,
      "nll": 1.7749,
      "ece": 0.2564
    },
    "workflow/jev_bench_tr/helpsteer2_verbosity": {
      "n": 1000,
      "accuracy": 0.299,
      "brier": 0.7382,
      "nll": 1.4194,
      "ece": 0.0526
    },
    "workflow/jev_bench_tr/ledgar": {
      "n": 1000,
      "accuracy": 0.089,
      "brier": 0.9987,
      "nll": 4.359,
      "ece": 0.1232
    },
    "workflow/jev_bench_tr/massive": {
      "n": 1000,
      "accuracy": 0.792,
      "brier": 0.3181,
      "nll": 0.9485,
      "ece": 0.0652
    },
    "workflow/jev_bench_tr/measuring_hate_speech": {
      "n": 1000,
      "accuracy": 0.602,
      "brier": 0.4646,
      "nll": 1.1807,
      "ece": 0.2498
    },
    "workflow/jev_bench_tr/mmlu": {
      "n": 1000,
      "accuracy": 0.258,
      "brier": 0.8901,
      "nll": 1.6694,
      "ece": 0.269
    },
    "workflow/jev_bench_tr/mnli": {
      "n": 1000,
      "accuracy": 0.769,
      "brier": 0.358,
      "nll": 0.6495,
      "ece": 0.078
    },
    "workflow/jev_bench_tr/paws": {
      "n": 1000,
      "accuracy": 0.461,
      "brier": 0.8029,
      "nll": 1.1797,
      "ece": 0.3917
    },
    "workflow/jev_bench_tr/sms_spam": {
      "n": 800,
      "accuracy": 0.955,
      "brier": 0.0716,
      "nll": 0.161,
      "ece": 0.0618
    },
    "workflow/jev_bench_tr/sst5": {
      "n": 1000,
      "accuracy": 0.374,
      "brier": 0.7501,
      "nll": 1.4187,
      "ece": 0.1727
    },
    "workflow/jev_bench_tr/strategyqa_closed": {
      "n": 687,
      "accuracy": 0.5124,
      "brier": 0.8002,
      "nll": 1.2016,
      "ece": 0.389
    },
    "workflow/jev_bench_tr/strategyqa_grounded": {
      "n": 687,
      "accuracy": 0.5473,
      "brier": 0.6726,
      "nll": 0.9917,
      "ece": 0.2883
    },
    "workflow/jev_bench_tr/stsb": {
      "n": 1000,
      "accuracy": 0.207,
      "brier": 0.8841,
      "nll": 1.838,
      "ece": 0.2022
    },
    "workflow/jev_bench_tr/yelp5": {
      "n": 1000,
      "accuracy": 0.436,
      "brier": 0.6977,
      "nll": 1.3257,
      "ece": 0.1277
    },
    "workflow/jev_bench_tr_en/arc_challenge": {
      "n": 91,
      "accuracy": 0.2967,
      "brier": 0.8757,
      "nll": 1.6404,
      "ece": 0.313
    },
    "workflow/jev_bench_tr_en/banking77": {
      "n": 71,
      "accuracy": 0.4648,
      "brier": 0.7657,
      "nll": 2.8336,
      "ece": 0.1895
    },
    "workflow/jev_bench_tr_en/boolq": {
      "n": 80,
      "accuracy": 0.725,
      "brier": 0.4328,
      "nll": 0.6599,
      "ece": 0.182
    },
    "workflow/jev_bench_tr_en/chaosnli": {
      "n": 151,
      "accuracy": 0.6821,
      "brier": 0.2307,
      "nll": 1.1925,
      "ece": 0.1487
    },
    "workflow/jev_bench_tr_en/civil_comments": {
      "n": 171,
      "accuracy": 0.883,
      "brier": 0.1086,
      "nll": 0.3648,
      "ece": 0.0575
    },
    "workflow/jev_bench_tr_en/clinc150": {
      "n": 96,
      "accuracy": 0.6042,
      "brier": 0.5402,
      "nll": 2.0147,
      "ece": 0.1222
    },
    "workflow/jev_bench_tr_en/fever_evidence": {
      "n": 91,
      "accuracy": 0.8242,
      "brier": 0.2424,
      "nll": 0.3865,
      "ece": 0.0749
    },
    "workflow/jev_bench_tr_en/go_emotions": {
      "n": 93,
      "accuracy": 0.4301,
      "brier": 0.3831,
      "nll": 2.703,
      "ece": 0.1576
    },
    "workflow/jev_bench_tr_en/helpsteer2_helpfulness": {
      "n": 89,
      "accuracy": 0.2697,
      "brier": 0.8004,
      "nll": 1.568,
      "ece": 0.1489
    },
    "workflow/jev_bench_tr_en/helpsteer2_verbosity": {
      "n": 79,
      "accuracy": 0.2532,
      "brier": 0.7294,
      "nll": 1.3758,
      "ece": 0.0877
    },
    "workflow/jev_bench_tr_en/ledgar": {
      "n": 70,
      "accuracy": 0.3286,
      "brier": 0.7949,
      "nll": 2.7906,
      "ece": 0.1533
    },
    "workflow/jev_bench_tr_en/massive": {
      "n": 95,
      "accuracy": 0.6947,
      "brier": 0.4596,
      "nll": 1.2966,
      "ece": 0.1514
    },
    "workflow/jev_bench_tr_en/measuring_hate_speech": {
      "n": 96,
      "accuracy": 0.5208,
      "brier": 0.4818,
      "nll": 1.1402,
      "ece": 0.1647
    },
    "workflow/jev_bench_tr_en/mmlu": {
      "n": 80,
      "accuracy": 0.3375,
      "brier": 0.8825,
      "nll": 1.6567,
      "ece": 0.2195
    },
    "workflow/jev_bench_tr_en/mnli": {
      "n": 92,
      "accuracy": 0.8804,
      "brier": 0.1863,
      "nll": 0.3435,
      "ece": 0.0482
    },
    "workflow/jev_bench_tr_en/paws": {
      "n": 80,
      "accuracy": 0.675,
      "brier": 0.3821,
      "nll": 0.5577,
      "ece": 0.1829
    },
    "workflow/jev_bench_tr_en/sms_spam": {
      "n": 82,
      "accuracy": 0.9878,
      "brier": 0.0332,
      "nll": 0.1108,
      "ece": 0.0644
    },
    "workflow/jev_bench_tr_en/sst5": {
      "n": 73,
      "accuracy": 0.4521,
      "brier": 0.6665,
      "nll": 1.1995,
      "ece": 0.2019
    },
    "workflow/jev_bench_tr_en/strategyqa_closed": {
      "n": 60,
      "accuracy": 0.55,
      "brier": 0.6608,
      "nll": 0.9526,
      "ece": 0.3191
    },
    "workflow/jev_bench_tr_en/strategyqa_grounded": {
      "n": 71,
      "accuracy": 0.6056,
      "brier": 0.5811,
      "nll": 0.8779,
      "ece": 0.2428
    },
    "workflow/jev_bench_tr_en/stsb": {
      "n": 90,
      "accuracy": 0.2444,
      "brier": 0.8279,
      "nll": 1.7187,
      "ece": 0.1456
    },
    "workflow/jev_bench_tr_en/yelp5": {
      "n": 99,
      "accuracy": 0.5354,
      "brier": 0.6303,
      "nll": 1.0981,
      "ece": 0.1693
    },
    "workflow/massive_tr": {
      "n": 2974,
      "accuracy": 0.7811,
      "brier": 0.3297,
      "nll": 0.9699,
      "ece": 0.0632
    },
    "workflow/xnli_tr": {
      "n": 5010,
      "accuracy": 0.7375,
      "brier": 0.3926,
      "nll": 0.6972,
      "ece": 0.0949
    }
  },
  "base": {
    "all": {
      "n": 89838,
      "accuracy": 0.4325,
      "brier": 0.7582,
      "nll": 2.0922,
      "ece": 0.277
    },
    "source/bev_tr": {
      "n": 46320,
      "accuracy": 0.4437,
      "brier": 0.7623,
      "nll": 2.0495,
      "ece": 0.2749
    },
    "source/bev_tr/choice": {
      "n": 18808,
      "accuracy": 0.4499,
      "brier": 0.7134,
      "nll": 1.5746,
      "ece": 0.2177
    },
    "source/bev_tr/noul": {
      "n": 17572,
      "accuracy": 0.5738,
      "brier": 0.6909,
      "nll": 2.5182,
      "ece": 0.2981
    },
    "source/bev_tr/score": {
      "n": 9940,
      "accuracy": 0.2021,
      "brier": 0.9807,
      "nll": 2.1195,
      "ece": 0.3422
    },
    "source/bev_tr_en": {
      "n": 5761,
      "accuracy": 0.4725,
      "brier": 0.7085,
      "nll": 1.8299,
      "ece": 0.2428
    },
    "source/bev_tr_en/choice": {
      "n": 2387,
      "accuracy": 0.4604,
      "brier": 0.6882,
      "nll": 1.5258,
      "ece": 0.2113
    },
    "source/bev_tr_en/noul": {
      "n": 2112,
      "accuracy": 0.6316,
      "brier": 0.5714,
      "nll": 1.8704,
      "ece": 0.2377
    },
    "source/bev_tr_en/score": {
      "n": 1262,
      "accuracy": 0.229,
      "brier": 0.9762,
      "nll": 2.3374,
      "ece": 0.3127
    },
    "source/beyazperde_tr": {
      "n": 5000,
      "accuracy": 0.1802,
      "brier": 0.9535,
      "nll": 1.9927,
      "ece": 0.3291
    },
    "source/beyazperde_tr/score": {
      "n": 5000,
      "accuracy": 0.1802,
      "brier": 0.9535,
      "nll": 1.9927,
      "ece": 0.3291
    },
    "source/jev_bench_tr": {
      "n": 22773,
      "accuracy": 0.402,
      "brier": 0.769,
      "nll": 2.3281,
      "ece": 0.2949
    },
    "source/jev_bench_tr/choice": {
      "n": 9599,
      "accuracy": 0.3403,
      "brier": 0.8114,
      "nll": 3.1437,
      "ece": 0.318
    },
    "source/jev_bench_tr/noul": {
      "n": 7174,
      "accuracy": 0.6391,
      "brier": 0.5953,
      "nll": 1.4894,
      "ece": 0.2635
    },
    "source/jev_bench_tr/score": {
      "n": 6000,
      "accuracy": 0.2172,
      "brier": 0.9089,
      "nll": 2.0262,
      "ece": 0.2957
    },
    "source/jev_bench_tr_en": {
      "n": 2000,
      "accuracy": 0.4995,
      "brier": 0.6504,
      "nll": 2.0143,
      "ece": 0.2458
    },
    "source/jev_bench_tr_en/choice": {
      "n": 839,
      "accuracy": 0.4803,
      "brier": 0.6591,
      "nll": 2.6277,
      "ece": 0.2565
    },
    "source/jev_bench_tr_en/noul": {
      "n": 635,
      "accuracy": 0.7213,
      "brier": 0.4063,
      "nll": 0.9013,
      "ece": 0.1565
    },
    "source/jev_bench_tr_en/score": {
      "n": 526,
      "accuracy": 0.2624,
      "brier": 0.9313,
      "nll": 2.3795,
      "ece": 0.3377
    },
    "source/massive_tr": {
      "n": 2974,
      "accuracy": 0.3547,
      "brier": 0.8699,
      "nll": 3.2285,
      "ece": 0.2849
    },
    "source/massive_tr/choice": {
      "n": 2974,
      "accuracy": 0.3547,
      "brier": 0.8699,
      "nll": 3.2285,
      "ece": 0.2849
    },
    "source/xnli_tr": {
      "n": 5010,
      "accuracy": 0.6926,
      "brier": 0.511,
      "nll": 1.1717,
      "ece": 0.2111
    },
    "source/xnli_tr/choice": {
      "n": 5010,
      "accuracy": 0.6926,
      "brier": 0.511,
      "nll": 1.1717,
      "ece": 0.2111
    },
    "type/choice": {
      "n": 39617,
      "accuracy": 0.4482,
      "brier": 0.7206,
      "nll": 2.0474,
      "ece": 0.2465
    },
    "type/noul": {
      "n": 27493,
      "accuracy": 0.5987,
      "brier": 0.6502,
      "nll": 2.1626,
      "ece": 0.2811
    },
    "type/score": {
      "n": 22728,
      "accuracy": 0.2042,
      "brier": 0.9544,
      "nll": 2.0851,
      "ece": 0.3252
    },
    "workflow/bev_tr/Browser interaction": {
      "n": 286,
      "accuracy": 0.2762,
      "brier": 0.8968,
      "nll": 1.9321,
      "ece": 0.2071
    },
    "workflow/bev_tr/Civic and safety operations": {
      "n": 261,
      "accuracy": 0.2529,
      "brier": 0.9125,
      "nll": 1.9973,
      "ece": 0.2501
    },
    "workflow/bev_tr/Contract evidence": {
      "n": 320,
      "accuracy": 0.3344,
      "brier": 0.9895,
      "nll": 1.7097,
      "ece": 0.4491
    },
    "workflow/bev_tr/Engagement and ranking": {
      "n": 82,
      "accuracy": 0.5488,
      "brier": 0.5359,
      "nll": 0.7447,
      "ece": 0.1386
    },
    "workflow/bev_tr/Financial reporting and banking": {
      "n": 1740,
      "accuracy": 0.1661,
      "brier": 1.2776,
      "nll": 8.0822,
      "ece": 0.5406
    },
    "workflow/bev_tr/Game-state decisions": {
      "n": 2039,
      "accuracy": 0.3448,
      "brier": 0.9951,
      "nll": 2.8801,
      "ece": 0.4139
    },
    "workflow/bev_tr/Reading comprehension": {
      "n": 1422,
      "accuracy": 0.6287,
      "brier": 0.5812,
      "nll": 1.0839,
      "ece": 0.2453
    },
    "workflow/bev_tr/Response preference and quality": {
      "n": 3176,
      "accuracy": 0.3237,
      "brier": 0.8783,
      "nll": 1.9154,
      "ece": 0.267
    },
    "workflow/bev_tr/Retail, product, and shopping": {
      "n": 6601,
      "accuracy": 0.2874,
      "brier": 1.0247,
      "nll": 3.8548,
      "ece": 0.4407
    },
    "workflow/bev_tr/Scientific and paper understanding": {
      "n": 508,
      "accuracy": 0.372,
      "brier": 0.9247,
      "nll": 1.8107,
      "ece": 0.4028
    },
    "workflow/bev_tr/Sensory quality rating": {
      "n": 260,
      "accuracy": 0.0731,
      "brier": 1.1058,
      "nll": 2.5668,
      "ece": 0.4822
    },
    "workflow/bev_tr/Sentiment, emotion, and moderation": {
      "n": 11526,
      "accuracy": 0.6164,
      "brier": 0.496,
      "nll": 0.896,
      "ece": 0.1549
    },
    "workflow/bev_tr/Software engineering and code": {
      "n": 2039,
      "accuracy": 0.3217,
      "brier": 0.7935,
      "nll": 1.5021,
      "ece": 0.1482
    },
    "workflow/bev_tr/Software security": {
      "n": 2240,
      "accuracy": 0.4147,
      "brier": 0.8559,
      "nll": 1.8811,
      "ece": 0.3557
    },
    "workflow/bev_tr/Spam detection": {
      "n": 271,
      "accuracy": 0.5498,
      "brier": 0.7614,
      "nll": 2.08,
      "ece": 0.3966
    },
    "workflow/bev_tr/Spatial and logical reasoning": {
      "n": 8384,
      "accuracy": 0.3204,
      "brier": 0.8964,
      "nll": 1.9159,
      "ece": 0.3173
    },
    "workflow/bev_tr/Support and intent routing": {
      "n": 2858,
      "accuracy": 0.6767,
      "brier": 0.4581,
      "nll": 0.9638,
      "ece": 0.1276
    },
    "workflow/bev_tr/Tool and workflow decisions": {
      "n": 2307,
      "accuracy": 0.7703,
      "brier": 0.3383,
      "nll": 0.7256,
      "ece": 0.1085
    },
    "workflow/bev_tr_en/Browser interaction": {
      "n": 39,
      "accuracy": 0.2821,
      "brier": 0.9354,
      "nll": 2.1869,
      "ece": 0.3387
    },
    "workflow/bev_tr_en/Civic and safety operations": {
      "n": 31,
      "accuracy": 0.0968,
      "brier": 1.0532,
      "nll": 2.4108,
      "ece": 0.4439
    },
    "workflow/bev_tr_en/Contract evidence": {
      "n": 49,
      "accuracy": 0.4286,
      "brier": 0.8485,
      "nll": 1.4634,
      "ece": 0.4012
    },
    "workflow/bev_tr_en/Engagement and ranking": {
      "n": 6,
      "accuracy": 0.6667,
      "brier": 0.4465,
      "nll": 0.6305,
      "ece": 0.2653
    },
    "workflow/bev_tr_en/Financial reporting and banking": {
      "n": 196,
      "accuracy": 0.1429,
      "brier": 1.297,
      "nll": 7.8218,
      "ece": 0.5676
    },
    "workflow/bev_tr_en/Game-state decisions": {
      "n": 303,
      "accuracy": 0.4983,
      "brier": 0.6867,
      "nll": 1.1958,
      "ece": 0.196
    },
    "workflow/bev_tr_en/Reading comprehension": {
      "n": 168,
      "accuracy": 0.6369,
      "brier": 0.5115,
      "nll": 0.8706,
      "ece": 0.2237
    },
    "workflow/bev_tr_en/Response preference and quality": {
      "n": 389,
      "accuracy": 0.3342,
      "brier": 0.8502,
      "nll": 1.7888,
      "ece": 0.2471
    },
    "workflow/bev_tr_en/Retail, product, and shopping": {
      "n": 902,
      "accuracy": 0.2971,
      "brier": 0.9608,
      "nll": 3.1178,
      "ece": 0.3787
    },
    "workflow/bev_tr_en/Scientific and paper understanding": {
      "n": 66,
      "accuracy": 0.3485,
      "brier": 1.012,
      "nll": 2.1604,
      "ece": 0.4553
    },
    "workflow/bev_tr_en/Sensory quality rating": {
      "n": 37,
      "accuracy": 0.1081,
      "brier": 0.8935,
      "nll": 1.85,
      "ece": 0.2321
    },
    "workflow/bev_tr_en/Sentiment, emotion, and moderation": {
      "n": 1433,
      "accuracy": 0.7139,
      "brier": 0.4116,
      "nll": 0.9539,
      "ece": 0.1375
    },
    "workflow/bev_tr_en/Software engineering and code": {
      "n": 260,
      "accuracy": 0.3231,
      "brier": 0.7653,
      "nll": 1.4255,
      "ece": 0.1383
    },
    "workflow/bev_tr_en/Software security": {
      "n": 234,
      "accuracy": 0.3718,
      "brier": 0.8652,
      "nll": 1.7216,
      "ece": 0.3089
    },
    "workflow/bev_tr_en/Spam detection": {
      "n": 23,
      "accuracy": 0.5652,
      "brier": 0.6446,
      "nll": 1.6813,
      "ece": 0.399
    },
    "workflow/bev_tr_en/Spatial and logical reasoning": {
      "n": 1042,
      "accuracy": 0.3081,
      "brier": 0.8798,
      "nll": 1.8558,
      "ece": 0.286
    },
    "workflow/bev_tr_en/Support and intent routing": {
      "n": 310,
      "accuracy": 0.7258,
      "brier": 0.3894,
      "nll": 0.8425,
      "ece": 0.1377
    },
    "workflow/bev_tr_en/Tool and workflow decisions": {
      "n": 273,
      "accuracy": 0.8022,
      "brier": 0.2895,
      "nll": 0.6295,
      "ece": 0.1321
    },
    "workflow/beyazperde_tr": {
      "n": 5000,
      "accuracy": 0.1802,
      "brier": 0.9535,
      "nll": 1.9927,
      "ece": 0.3291
    },
    "workflow/jev_bench_tr/arc_challenge": {
      "n": 1000,
      "accuracy": 0.249,
      "brier": 0.8758,
      "nll": 1.6658,
      "ece": 0.2392
    },
    "workflow/jev_bench_tr/banking77": {
      "n": 1000,
      "accuracy": 0.27,
      "brier": 1.1483,
      "nll": 5.3527,
      "ece": 0.4696
    },
    "workflow/jev_bench_tr/boolq": {
      "n": 1000,
      "accuracy": 0.687,
      "brier": 0.5095,
      "nll": 1.0117,
      "ece": 0.2262
    },
    "workflow/jev_bench_tr/chaosnli": {
      "n": 1599,
      "accuracy": 0.4772,
      "brier": 0.4824,
      "nll": 2.3205,
      "ece": 0.3991
    },
    "workflow/jev_bench_tr/civil_comments": {
      "n": 2000,
      "accuracy": 0.9125,
      "brier": 0.0821,
      "nll": 0.3281,
      "ece": 0.0263
    },
    "workflow/jev_bench_tr/clinc150": {
      "n": 1000,
      "accuracy": 0.25,
      "brier": 0.9822,
      "nll": 4.0163,
      "ece": 0.2849
    },
    "workflow/jev_bench_tr/fever_evidence": {
      "n": 1000,
      "accuracy": 0.551,
      "brier": 0.7461,
      "nll": 1.4558,
      "ece": 0.3511
    },
    "workflow/jev_bench_tr/go_emotions": {
      "n": 1000,
      "accuracy": 0.278,
      "brier": 0.73,
      "nll": 4.5711,
      "ece": 0.4802
    },
    "workflow/jev_bench_tr/helpsteer2_helpfulness": {
      "n": 1000,
      "accuracy": 0.15,
      "brier": 0.9073,
      "nll": 1.8959,
      "ece": 0.2547
    },
    "workflow/jev_bench_tr/helpsteer2_verbosity": {
      "n": 1000,
      "accuracy": 0.211,
      "brier": 0.8589,
      "nll": 1.6432,
      "ece": 0.2394
    },
    "workflow/jev_bench_tr/ledgar": {
      "n": 1000,
      "accuracy": 0.087,
      "brier": 1.1279,
      "nll": 5.1478,
      "ece": 0.2658
    },
    "workflow/jev_bench_tr/massive": {
      "n": 1000,
      "accuracy": 0.375,
      "brier": 0.8405,
      "nll": 3.0528,
      "ece": 0.2758
    },
    "workflow/jev_bench_tr/measuring_hate_speech": {
      "n": 1000,
      "accuracy": 0.313,
      "brier": 0.5368,
      "nll": 1.2331,
      "ece": 0.1656
    },
    "workflow/jev_bench_tr/mmlu": {
      "n": 1000,
      "accuracy": 0.275,
      "brier": 0.8531,
      "nll": 1.6323,
      "ece": 0.2134
    },
    "workflow/jev_bench_tr/mnli": {
      "n": 1000,
      "accuracy": 0.72,
      "brier": 0.459,
      "nll": 1.0265,
      "ece": 0.1898
    },
    "workflow/jev_bench_tr/paws": {
      "n": 1000,
      "accuracy": 0.528,
      "brier": 0.8257,
      "nll": 1.7868,
      "ece": 0.402
    },
    "workflow/jev_bench_tr/sms_spam": {
      "n": 800,
      "accuracy": 0.3275,
      "brier": 1.2108,
      "nll": 4.8477,
      "ece": 0.6043
    },
    "workflow/jev_bench_tr/sst5": {
      "n": 1000,
      "accuracy": 0.286,
      "brier": 0.8187,
      "nll": 1.6116,
      "ece": 0.1613
    },
    "workflow/jev_bench_tr/strategyqa_closed": {
      "n": 687,
      "accuracy": 0.5167,
      "brier": 0.8117,
      "nll": 1.4146,
      "ece": 0.378
    },
    "workflow/jev_bench_tr/strategyqa_grounded": {
      "n": 687,
      "accuracy": 0.5488,
      "brier": 0.726,
      "nll": 1.3457,
      "ece": 0.3305
    },
    "workflow/jev_bench_tr/stsb": {
      "n": 1000,
      "accuracy": 0.158,
      "brier": 1.1593,
      "nll": 2.5632,
      "ece": 0.4602
    },
    "workflow/jev_bench_tr/yelp5": {
      "n": 1000,
      "accuracy": 0.185,
      "brier": 1.1724,
      "nll": 3.2104,
      "ece": 0.4975
    },
    "workflow/jev_bench_tr_en/arc_challenge": {
      "n": 91,
      "accuracy": 0.3077,
      "brier": 0.8521,
      "nll": 1.618,
      "ece": 0.2606
    },
    "workflow/jev_bench_tr_en/banking77": {
      "n": 71,
      "accuracy": 0.4366,
      "brier": 0.9091,
      "nll": 4.9136,
      "ece": 0.4145
    },
    "workflow/jev_bench_tr_en/boolq": {
      "n": 80,
      "accuracy": 0.725,
      "brier": 0.4529,
      "nll": 0.886,
      "ece": 0.219
    },
    "workflow/jev_bench_tr_en/chaosnli": {
      "n": 151,
      "accuracy": 0.6225,
      "brier": 0.3674,
      "nll": 1.9774,
      "ece": 0.2756
    },
    "workflow/jev_bench_tr_en/civil_comments": {
      "n": 171,
      "accuracy": 0.9532,
      "brier": 0.0364,
      "nll": 0.253,
      "ece": 0.0844
    },
    "workflow/jev_bench_tr_en/clinc150": {
      "n": 96,
      "accuracy": 0.4792,
      "brier": 0.7581,
      "nll": 2.9071,
      "ece": 0.309
    },
    "workflow/jev_bench_tr_en/fever_evidence": {
      "n": 91,
      "accuracy": 0.7253,
      "brier": 0.4523,
      "nll": 0.7521,
      "ece": 0.2241
    },
    "workflow/jev_bench_tr_en/go_emotions": {
      "n": 93,
      "accuracy": 0.3978,
      "brier": 0.5766,
      "nll": 4.0598,
      "ece": 0.375
    },
    "workflow/jev_bench_tr_en/helpsteer2_helpfulness": {
      "n": 89,
      "accuracy": 0.1348,
      "brier": 0.8934,
      "nll": 1.7886,
      "ece": 0.3137
    },
    "workflow/jev_bench_tr_en/helpsteer2_verbosity": {
      "n": 79,
      "accuracy": 0.1646,
      "brier": 0.9505,
      "nll": 1.6857,
      "ece": 0.3926
    },
    "workflow/jev_bench_tr_en/ledgar": {
      "n": 70,
      "accuracy": 0.2714,
      "brier": 0.9955,
      "nll": 4.4382,
      "ece": 0.32
    },
    "workflow/jev_bench_tr_en/massive": {
      "n": 95,
      "accuracy": 0.4105,
      "brier": 0.8485,
      "nll": 3.043,
      "ece": 0.2822
    },
    "workflow/jev_bench_tr_en/measuring_hate_speech": {
      "n": 96,
      "accuracy": 0.4479,
      "brier": 0.482,
      "nll": 1.085,
      "ece": 0.074
    },
    "workflow/jev_bench_tr_en/mmlu": {
      "n": 80,
      "accuracy": 0.3375,
      "brier": 0.7891,
      "nll": 1.4972,
      "ece": 0.1576
    },
    "workflow/jev_bench_tr_en/mnli": {
      "n": 92,
      "accuracy": 0.8913,
      "brier": 0.1693,
      "nll": 0.3673,
      "ece": 0.0821
    },
    "workflow/jev_bench_tr_en/paws": {
      "n": 80,
      "accuracy": 0.65,
      "brier": 0.4374,
      "nll": 0.6762,
      "ece": 0.2444
    },
    "workflow/jev_bench_tr_en/sms_spam": {
      "n": 82,
      "accuracy": 0.5488,
      "brier": 0.7829,
      "nll": 2.661,
      "ece": 0.3921
    },
    "workflow/jev_bench_tr_en/sst5": {
      "n": 73,
      "accuracy": 0.3014,
      "brier": 1.0207,
      "nll": 2.3759,
      "ece": 0.4117
    },
    "workflow/jev_bench_tr_en/strategyqa_closed": {
      "n": 60,
      "accuracy": 0.5333,
      "brier": 0.5838,
      "nll": 0.8489,
      "ece": 0.2998
    },
    "workflow/jev_bench_tr_en/strategyqa_grounded": {
      "n": 71,
      "accuracy": 0.5915,
      "brier": 0.5658,
      "nll": 0.9366,
      "ece": 0.2513
    },
    "workflow/jev_bench_tr_en/stsb": {
      "n": 90,
      "accuracy": 0.2333,
      "brier": 1.0749,
      "nll": 2.5195,
      "ece": 0.3886
    },
    "workflow/jev_bench_tr_en/yelp5": {
      "n": 99,
      "accuracy": 0.2727,
      "brier": 1.1895,
      "nll": 4.5951,
      "ece": 0.5002
    },
    "workflow/massive_tr": {
      "n": 2974,
      "accuracy": 0.3547,
      "brier": 0.8699,
      "nll": 3.2285,
      "ece": 0.2849
    },
    "workflow/xnli_tr": {
      "n": 5010,
      "accuracy": 0.6926,
      "brier": 0.511,
      "nll": 1.1717,
      "ece": 0.2111
    }
  }
}