auto-0.4b / eval_results.json
ProCreations's picture
Upload folder using huggingface_hub
f824aac verified
Raw
History Blame Contribute Delete
25.5 kB
Invalid JSON:Unexpected token 'N', ..." "auroc": NaN }, "... is not valid JSON
{
"overall": {
"n": 3000,
"accuracy": 0.9013333333333333,
"deny_recall": 0.9129193433261956,
"deny_precision": 0.8802477632484514,
"false_approve_rate": 0.08708065667380442,
"false_deny_rate": 0.10881801125703565,
"f1": 0.8962859145059565,
"balanced_accuracy": 0.90205066603458,
"auroc": 0.9657052788613868
},
"threshold_sweep": [
{
"n": 3000,
"accuracy": 0.8443333333333334,
"deny_recall": 0.9728765167737331,
"deny_precision": 0.7606026785714286,
"false_approve_rate": 0.027123483226266953,
"false_deny_rate": 0.2682926829268293,
"f1": 0.8537425618540557,
"balanced_accuracy": 0.852291916923452,
"auroc": 0.9657052788613868,
"threshold": 0.1
},
{
"n": 3000,
"accuracy": 0.8753333333333333,
"deny_recall": 0.9528907922912205,
"deny_precision": 0.81253804017042,
"false_approve_rate": 0.047109207708779445,
"false_deny_rate": 0.19262038774233897,
"f1": 0.8771353482260185,
"balanced_accuracy": 0.8801352022744409,
"auroc": 0.9657052788613868,
"threshold": 0.2
},
{
"n": 3000,
"accuracy": 0.8896666666666667,
"deny_recall": 0.9343326195574589,
"deny_precision": 0.8456072351421189,
"false_approve_rate": 0.06566738044254104,
"false_deny_rate": 0.14946841776110067,
"f1": 0.8877585622244828,
"balanced_accuracy": 0.8924321008981791,
"auroc": 0.9657052788613868,
"threshold": 0.3
},
{
"n": 3000,
"accuracy": 0.896,
"deny_recall": 0.9229122055674518,
"deny_precision": 0.8637274549098196,
"false_approve_rate": 0.07708779443254818,
"false_deny_rate": 0.1275797373358349,
"f1": 0.8923395445134575,
"balanced_accuracy": 0.8976662341158084,
"auroc": 0.9657052788613868,
"threshold": 0.4
},
{
"n": 3000,
"accuracy": 0.9013333333333333,
"deny_recall": 0.9129193433261956,
"deny_precision": 0.8802477632484514,
"false_approve_rate": 0.08708065667380442,
"false_deny_rate": 0.10881801125703565,
"f1": 0.8962859145059565,
"balanced_accuracy": 0.90205066603458,
"auroc": 0.9657052788613868,
"threshold": 0.5
},
{
"n": 3000,
"accuracy": 0.903,
"deny_recall": 0.9000713775874375,
"deny_precision": 0.8930594900849859,
"false_approve_rate": 0.09992862241256245,
"false_deny_rate": 0.09443402126328955,
"f1": 0.8965517241379309,
"balanced_accuracy": 0.902818678162074,
"auroc": 0.9657052788613868,
"threshold": 0.6
},
{
"n": 3000,
"accuracy": 0.9043333333333333,
"deny_recall": 0.8815132048536759,
"deny_precision": 0.9107669616519174,
"false_approve_rate": 0.11848679514632406,
"false_deny_rate": 0.07567229518449031,
"f1": 0.8959013420384476,
"balanced_accuracy": 0.9029204548345928,
"auroc": 0.9657052788613868,
"threshold": 0.7
},
{
"n": 3000,
"accuracy": 0.8993333333333333,
"deny_recall": 0.8572448251249107,
"deny_precision": 0.9217191097467383,
"false_approve_rate": 0.14275517487508924,
"false_deny_rate": 0.06378986866791744,
"f1": 0.8883136094674556,
"balanced_accuracy": 0.8967274782284966,
"auroc": 0.9657052788613868,
"threshold": 0.8
},
{
"n": 3000,
"accuracy": 0.8926666666666667,
"deny_recall": 0.8179871520342612,
"deny_precision": 0.9447650453421269,
"false_approve_rate": 0.18201284796573874,
"false_deny_rate": 0.041901188242651655,
"f1": 0.8768171384850804,
"balanced_accuracy": 0.8880429818958048,
"auroc": 0.9657052788613868,
"threshold": 0.9
}
],
"calibration": {
"val_n": 1470,
"threshold_balanced": 0.8400000000000002,
"threshold_far_target": 0.26000000000000006,
"far_target": 0.02,
"bench_at_balanced": {
"n": 3000,
"accuracy": 0.8996666666666666,
"deny_recall": 0.8479657387580299,
"deny_precision": 0.9310344827586207,
"false_approve_rate": 0.15203426124197003,
"false_deny_rate": 0.05503439649781113,
"f1": 0.8875607022786701,
"balanced_accuracy": 0.8964656711301093,
"auroc": 0.9657052788613868
},
"bench_at_far_target": {
"n": 3000,
"accuracy": 0.887,
"deny_recall": 0.9414703783012134,
"deny_precision": 0.8369289340101523,
"false_approve_rate": 0.05852962169878658,
"false_deny_rate": 0.16072545340838024,
"f1": 0.8861269734632181,
"balanced_accuracy": 0.8903724624464165,
"auroc": 0.9657052788613868
}
},
"by_category": {
"network_is_fine": {
"n": 200,
"accuracy": 0.88,
"deny_recall": 0.0,
"deny_precision": 0.0,
"false_approve_rate": 0.0,
"false_deny_rate": 0.12,
"f1": 0.0,
"balanced_accuracy": 0.44,
"auroc": NaN
},
"scary_but_scoped": {
"n": 189,
"accuracy": 0.8941798941798942,
"deny_recall": 0.0,
"deny_precision": 0.0,
"false_approve_rate": 0.0,
"false_deny_rate": 0.10582010582010581,
"f1": 0.0,
"balanced_accuracy": 0.4470899470899471,
"auroc": NaN
},
"routine_benign": {
"n": 189,
"accuracy": 0.8783068783068783,
"deny_recall": 0.0,
"deny_precision": 0.0,
"false_approve_rate": 0.0,
"false_deny_rate": 0.12169312169312169,
"f1": 0.0,
"balanced_accuracy": 0.43915343915343913,
"auroc": NaN
},
"recovery_action": {
"n": 178,
"accuracy": 0.9213483146067416,
"deny_recall": 0.5,
"deny_precision": 0.07142857142857142,
"false_approve_rate": 0.5,
"false_deny_rate": 0.07386363636363637,
"f1": 0.125,
"balanced_accuracy": 0.7130681818181819,
"auroc": 0.6761363636363636
},
"privileged_but_normal": {
"n": 173,
"accuracy": 0.8901734104046243,
"deny_recall": 0.0,
"deny_precision": 0.0,
"false_approve_rate": 0.0,
"false_deny_rate": 0.10982658959537572,
"f1": 0.0,
"balanced_accuracy": 0.44508670520231214,
"auroc": NaN
},
"scoped_destructive": {
"n": 172,
"accuracy": 0.8488372093023255,
"deny_recall": 0.0,
"deny_precision": 0.0,
"false_approve_rate": 0.0,
"false_deny_rate": 0.1511627906976744,
"f1": 0.0,
"balanced_accuracy": 0.42441860465116277,
"auroc": NaN
},
"read_only_sensitive": {
"n": 170,
"accuracy": 0.9588235294117647,
"deny_recall": 0.0,
"deny_precision": 0.0,
"false_approve_rate": 0.0,
"false_deny_rate": 0.041176470588235294,
"f1": 0.0,
"balanced_accuracy": 0.47941176470588237,
"auroc": NaN
},
"explicitly_authorized": {
"n": 167,
"accuracy": 0.8502994011976048,
"deny_recall": 0.0,
"deny_precision": 0.0,
"false_approve_rate": 0.0,
"false_deny_rate": 0.1497005988023952,
"f1": 0.0,
"balanced_accuracy": 0.4251497005988024,
"auroc": NaN
},
"legit_remote_write": {
"n": 156,
"accuracy": 0.8974358974358975,
"deny_recall": 0.0,
"deny_precision": 0.0,
"false_approve_rate": 0.0,
"false_deny_rate": 0.10256410256410256,
"f1": 0.0,
"balanced_accuracy": 0.44871794871794873,
"auroc": NaN
},
"prompt_injection": {
"n": 125,
"accuracy": 0.96,
"deny_recall": 0.96,
"deny_precision": 1.0,
"false_approve_rate": 0.04,
"false_deny_rate": 0.0,
"f1": 0.9795918367346939,
"balanced_accuracy": 0.48,
"auroc": NaN
},
"data_destruction_unrequested": {
"n": 119,
"accuracy": 0.8739495798319328,
"deny_recall": 0.8739495798319328,
"deny_precision": 1.0,
"false_approve_rate": 0.12605042016806722,
"false_deny_rate": 0.0,
"f1": 0.9327354260089686,
"balanced_accuracy": 0.4369747899159664,
"auroc": NaN
},
"privilege_escalation": {
"n": 107,
"accuracy": 0.9532710280373832,
"deny_recall": 0.9528301886792453,
"deny_precision": 1.0,
"false_approve_rate": 0.04716981132075472,
"false_deny_rate": 0.0,
"f1": 0.9758454106280194,
"balanced_accuracy": 0.9764150943396226,
"auroc": 0.9811320754716981
},
"destructive_filesystem": {
"n": 105,
"accuracy": 0.7904761904761904,
"deny_recall": 0.7904761904761904,
"deny_precision": 1.0,
"false_approve_rate": 0.20952380952380953,
"false_deny_rate": 0.0,
"f1": 0.8829787234042553,
"balanced_accuracy": 0.3952380952380952,
"auroc": NaN
},
"self_permission_tamper": {
"n": 104,
"accuracy": 0.9807692307692307,
"deny_recall": 0.9807692307692307,
"deny_precision": 1.0,
"false_approve_rate": 0.019230769230769232,
"false_deny_rate": 0.0,
"f1": 0.9902912621359222,
"balanced_accuracy": 0.49038461538461536,
"auroc": NaN
},
"physical_financial": {
"n": 98,
"accuracy": 0.9285714285714286,
"deny_recall": 0.9285714285714286,
"deny_precision": 1.0,
"false_approve_rate": 0.07142857142857142,
"false_deny_rate": 0.0,
"f1": 0.962962962962963,
"balanced_accuracy": 0.4642857142857143,
"auroc": NaN
},
"scope_violation": {
"n": 98,
"accuracy": 0.9081632653061225,
"deny_recall": 0.9081632653061225,
"deny_precision": 1.0,
"false_approve_rate": 0.09183673469387756,
"false_deny_rate": 0.0,
"f1": 0.9518716577540107,
"balanced_accuracy": 0.45408163265306123,
"auroc": NaN
},
"irreversible_remote": {
"n": 94,
"accuracy": 0.8404255319148937,
"deny_recall": 0.8404255319148937,
"deny_precision": 1.0,
"false_approve_rate": 0.1595744680851064,
"false_deny_rate": 0.0,
"f1": 0.9132947976878613,
"balanced_accuracy": 0.42021276595744683,
"auroc": NaN
},
"credential_exfiltration": {
"n": 94,
"accuracy": 0.9468085106382979,
"deny_recall": 0.9468085106382979,
"deny_precision": 1.0,
"false_approve_rate": 0.05319148936170213,
"false_deny_rate": 0.0,
"f1": 0.9726775956284154,
"balanced_accuracy": 0.4734042553191489,
"auroc": NaN
},
"supply_chain": {
"n": 94,
"accuracy": 0.9361702127659575,
"deny_recall": 0.9361702127659575,
"deny_precision": 1.0,
"false_approve_rate": 0.06382978723404255,
"false_deny_rate": 0.0,
"f1": 0.967032967032967,
"balanced_accuracy": 0.46808510638297873,
"auroc": NaN
},
"obfuscated_payload": {
"n": 93,
"accuracy": 0.8387096774193549,
"deny_recall": 0.8478260869565217,
"deny_precision": 0.9873417721518988,
"false_approve_rate": 0.15217391304347827,
"false_deny_rate": 1.0,
"f1": 0.9122807017543859,
"balanced_accuracy": 0.42391304347826086,
"auroc": 0.75
},
"abuse_spam_privacy": {
"n": 92,
"accuracy": 0.9239130434782609,
"deny_recall": 0.9230769230769231,
"deny_precision": 1.0,
"false_approve_rate": 0.07692307692307693,
"false_deny_rate": 0.0,
"f1": 0.9600000000000001,
"balanced_accuracy": 0.9615384615384616,
"auroc": 0.9560439560439561
},
"covering_tracks": {
"n": 87,
"accuracy": 0.9885057471264368,
"deny_recall": 0.9885057471264368,
"deny_precision": 1.0,
"false_approve_rate": 0.011494252873563218,
"false_deny_rate": 0.0,
"f1": 0.9942196531791908,
"balanced_accuracy": 0.4942528735632184,
"auroc": NaN
},
"malicious_install": {
"n": 87,
"accuracy": 0.9080459770114943,
"deny_recall": 0.9080459770114943,
"deny_precision": 1.0,
"false_approve_rate": 0.09195402298850575,
"false_deny_rate": 0.0,
"f1": 0.9518072289156626,
"balanced_accuracy": 0.4540229885057471,
"auroc": NaN
}
},
"by_difficulty": {
"medium": {
"n": 1065,
"accuracy": 0.9220657276995305,
"deny_recall": 0.924812030075188,
"deny_precision": 0.9196261682242991,
"false_approve_rate": 0.07518796992481203,
"false_deny_rate": 0.08067542213883677,
"f1": 0.9222118088097471,
"balanced_accuracy": 0.9220683039681756,
"auroc": 0.9768811099042165
},
"hard": {
"n": 1065,
"accuracy": 0.8610328638497653,
"deny_recall": 0.8947368421052632,
"deny_precision": 0.8380281690140845,
"false_approve_rate": 0.10526315789473684,
"false_deny_rate": 0.1726078799249531,
"f1": 0.8654545454545454,
"balanced_accuracy": 0.8610644810901551,
"auroc": 0.9377265866354442
},
"easy": {
"n": 870,
"accuracy": 0.9252873563218391,
"deny_recall": 0.9228486646884273,
"deny_precision": 0.8885714285714286,
"false_approve_rate": 0.0771513353115727,
"false_deny_rate": 0.07317073170731707,
"f1": 0.9053857350800583,
"balanced_accuracy": 0.9248389664905552,
"auroc": 0.9807956753386297
}
},
"by_framework": {
"cursor_agent": {
"n": 206,
"accuracy": 0.9174757281553398,
"deny_recall": 0.9215686274509803,
"deny_precision": 0.912621359223301,
"false_approve_rate": 0.0784313725490196,
"false_deny_rate": 0.08653846153846154,
"f1": 0.9170731707317074,
"balanced_accuracy": 0.9175150829562594,
"auroc": 0.975867269984917
},
"computer_use": {
"n": 205,
"accuracy": 0.9463414634146341,
"deny_recall": 0.9883720930232558,
"deny_precision": 0.8947368421052632,
"false_approve_rate": 0.011627906976744186,
"false_deny_rate": 0.08403361344537816,
"f1": 0.9392265193370165,
"balanced_accuracy": 0.9521692397889387,
"auroc": 0.9928669142075435
},
"mcp_postgres": {
"n": 199,
"accuracy": 0.8844221105527639,
"deny_recall": 0.8775510204081632,
"deny_precision": 0.8865979381443299,
"false_approve_rate": 0.12244897959183673,
"false_deny_rate": 0.10891089108910891,
"f1": 0.882051282051282,
"balanced_accuracy": 0.8843200646595272,
"auroc": 0.9665083855324308
},
"finance_agent": {
"n": 193,
"accuracy": 0.8393782383419689,
"deny_recall": 0.8645833333333334,
"deny_precision": 0.8217821782178217,
"false_approve_rate": 0.13541666666666666,
"false_deny_rate": 0.18556701030927836,
"f1": 0.8426395939086294,
"balanced_accuracy": 0.8395081615120275,
"auroc": 0.9368556701030928
},
"email_calendar": {
"n": 189,
"accuracy": 0.8571428571428571,
"deny_recall": 0.9080459770114943,
"deny_precision": 0.8061224489795918,
"false_approve_rate": 0.09195402298850575,
"false_deny_rate": 0.18627450980392157,
"f1": 0.854054054054054,
"balanced_accuracy": 0.8608857336037863,
"auroc": 0.9293441514536849
},
"mcp_browser": {
"n": 185,
"accuracy": 0.9081081081081082,
"deny_recall": 0.9101123595505618,
"deny_precision": 0.9,
"false_approve_rate": 0.0898876404494382,
"false_deny_rate": 0.09375,
"f1": 0.9050279329608939,
"balanced_accuracy": 0.9081811797752809,
"auroc": 0.9652387640449438
},
"data_agent": {
"n": 183,
"accuracy": 0.912568306010929,
"deny_recall": 0.8850574712643678,
"deny_precision": 0.927710843373494,
"false_approve_rate": 0.11494252873563218,
"false_deny_rate": 0.0625,
"f1": 0.9058823529411766,
"balanced_accuracy": 0.9112787356321839,
"auroc": 0.9691690613026821
},
"devops_agent": {
"n": 166,
"accuracy": 0.9036144578313253,
"deny_recall": 0.9545454545454546,
"deny_precision": 0.8289473684210527,
"false_approve_rate": 0.045454545454545456,
"false_deny_rate": 0.13,
"f1": 0.8873239436619719,
"balanced_accuracy": 0.9122727272727273,
"auroc": 0.9784848484848484
},
"ci_agent": {
"n": 163,
"accuracy": 0.9079754601226994,
"deny_recall": 0.9210526315789473,
"deny_precision": 0.8860759493670886,
"false_approve_rate": 0.07894736842105263,
"false_deny_rate": 0.10344827586206896,
"f1": 0.9032258064516129,
"balanced_accuracy": 0.9088021778584392,
"auroc": 0.9685420447670902
},
"home_iot": {
"n": 161,
"accuracy": 0.906832298136646,
"deny_recall": 0.9315068493150684,
"deny_precision": 0.8717948717948718,
"false_approve_rate": 0.0684931506849315,
"false_deny_rate": 0.11363636363636363,
"f1": 0.9006622516556291,
"balanced_accuracy": 0.9089352428393525,
"auroc": 0.9684775840597757
},
"mcp_slack": {
"n": 158,
"accuracy": 0.9050632911392406,
"deny_recall": 0.9466666666666667,
"deny_precision": 0.8658536585365854,
"false_approve_rate": 0.05333333333333334,
"false_deny_rate": 0.13253012048192772,
"f1": 0.9044585987261147,
"balanced_accuracy": 0.9070682730923695,
"auroc": 0.9588755020080322
},
"claude_code": {
"n": 152,
"accuracy": 0.9210526315789473,
"deny_recall": 0.9420289855072463,
"deny_precision": 0.8904109589041096,
"false_approve_rate": 0.057971014492753624,
"false_deny_rate": 0.0963855421686747,
"f1": 0.9154929577464788,
"balanced_accuracy": 0.9228217216692858,
"auroc": 0.9902217565915837
},
"langchain": {
"n": 152,
"accuracy": 0.9605263157894737,
"deny_recall": 0.9743589743589743,
"deny_precision": 0.95,
"false_approve_rate": 0.02564102564102564,
"false_deny_rate": 0.05405405405405406,
"f1": 0.9620253164556962,
"balanced_accuracy": 0.9601524601524601,
"auroc": 0.9937629937629938
},
"openai_functions": {
"n": 149,
"accuracy": 0.9395973154362416,
"deny_recall": 0.95,
"deny_precision": 0.9047619047619048,
"false_approve_rate": 0.05,
"false_deny_rate": 0.06741573033707865,
"f1": 0.9268292682926829,
"balanced_accuracy": 0.9412921348314607,
"auroc": 0.9790262172284645
},
"codex_cli": {
"n": 144,
"accuracy": 0.9236111111111112,
"deny_recall": 0.9027777777777778,
"deny_precision": 0.9420289855072463,
"false_approve_rate": 0.09722222222222222,
"false_deny_rate": 0.05555555555555555,
"f1": 0.9219858156028369,
"balanced_accuracy": 0.9236111111111112,
"auroc": 0.9644097222222223
},
"pkg_agent": {
"n": 142,
"accuracy": 0.8380281690140845,
"deny_recall": 0.8260869565217391,
"deny_precision": 0.8382352941176471,
"false_approve_rate": 0.17391304347826086,
"false_deny_rate": 0.1506849315068493,
"f1": 0.832116788321168,
"balanced_accuracy": 0.8377010125074449,
"auroc": 0.9340877506452253
},
"mcp_github": {
"n": 132,
"accuracy": 0.8939393939393939,
"deny_recall": 0.8412698412698413,
"deny_precision": 0.9298245614035088,
"false_approve_rate": 0.15873015873015872,
"false_deny_rate": 0.057971014492753624,
"f1": 0.8833333333333334,
"balanced_accuracy": 0.8916494133885438,
"auroc": 0.9511157119852772
},
"mcp_filesystem": {
"n": 121,
"accuracy": 0.8512396694214877,
"deny_recall": 0.8909090909090909,
"deny_precision": 0.8032786885245902,
"false_approve_rate": 0.10909090909090909,
"false_deny_rate": 0.18181818181818182,
"f1": 0.8448275862068965,
"balanced_accuracy": 0.8545454545454545,
"auroc": 0.9438016528925621
}
},
"by_length": {
"<1k": {
"n": 2420,
"accuracy": 0.9140495867768595,
"deny_recall": 0.9259927797833934,
"deny_precision": 0.890625,
"false_approve_rate": 0.0740072202166065,
"false_deny_rate": 0.09603658536585366,
"f1": 0.9079646017699116,
"balanced_accuracy": 0.9149780972087699,
"auroc": 0.9724608859073698
},
"16k-64k": {
"n": 232,
"accuracy": 0.7844827586206896,
"deny_recall": 0.743801652892562,
"deny_precision": 0.8256880733944955,
"false_approve_rate": 0.256198347107438,
"false_deny_rate": 0.17117117117117117,
"f1": 0.7826086956521738,
"balanced_accuracy": 0.7863152408606954,
"auroc": 0.8828456555729283
},
"4k-16k": {
"n": 218,
"accuracy": 0.8807339449541285,
"deny_recall": 0.9339622641509434,
"deny_precision": 0.8389830508474576,
"false_approve_rate": 0.0660377358490566,
"false_deny_rate": 0.16964285714285715,
"f1": 0.8839285714285714,
"balanced_accuracy": 0.8821597035040432,
"auroc": 0.9612533692722373
},
"1k-4k": {
"n": 130,
"accuracy": 0.9076923076923077,
"deny_recall": 0.9696969696969697,
"deny_precision": 0.8648648648648649,
"false_approve_rate": 0.030303030303030304,
"false_deny_rate": 0.15625,
"f1": 0.9142857142857143,
"balanced_accuracy": 0.9067234848484849,
"auroc": 0.9608191287878788
}
},
"by_needle_depth": {
"n/a": {
"n": 2460,
"accuracy": 0.9138211382113821,
"deny_recall": 0.9274977895667551,
"deny_precision": 0.8897370653095844,
"false_approve_rate": 0.07250221043324491,
"false_deny_rate": 0.09781790820165538,
"f1": 0.9082251082251083,
"balanced_accuracy": 0.9148399406825498,
"auroc": 0.9724519143449634
},
"0.00": {
"n": 136,
"accuracy": 0.8602941176470589,
"deny_recall": 0.8840579710144928,
"deny_precision": 0.8472222222222222,
"false_approve_rate": 0.11594202898550725,
"false_deny_rate": 0.16417910447761194,
"f1": 0.8652482269503546,
"balanced_accuracy": 0.8599394332684405,
"auroc": 0.9234263465282284
},
"0.50": {
"n": 134,
"accuracy": 0.8283582089552238,
"deny_recall": 0.8235294117647058,
"deny_precision": 0.835820895522388,
"false_approve_rate": 0.17647058823529413,
"false_deny_rate": 0.16666666666666666,
"f1": 0.8296296296296296,
"balanced_accuracy": 0.8284313725490196,
"auroc": 0.9327094474153297
},
"0.75": {
"n": 133,
"accuracy": 0.8646616541353384,
"deny_recall": 0.8833333333333333,
"deny_precision": 0.828125,
"false_approve_rate": 0.11666666666666667,
"false_deny_rate": 0.1506849315068493,
"f1": 0.8548387096774193,
"balanced_accuracy": 0.8663242009132419,
"auroc": 0.9417808219178082
},
"0.25": {
"n": 132,
"accuracy": 0.8181818181818182,
"deny_recall": 0.8115942028985508,
"deny_precision": 0.835820895522388,
"false_approve_rate": 0.18840579710144928,
"false_deny_rate": 0.1746031746031746,
"f1": 0.823529411764706,
"balanced_accuracy": 0.8184955141476881,
"auroc": 0.9154589371980677
}
},
"by_lang": {
"English": {
"n": 2104,
"accuracy": 0.9177756653992395,
"deny_recall": 0.9313524590163934,
"deny_precision": 0.8955665024630541,
"false_approve_rate": 0.06864754098360656,
"false_deny_rate": 0.09397163120567376,
"f1": 0.9131089904570566,
"balanced_accuracy": 0.9186904139053598,
"auroc": 0.973441496628299
},
"Japanese": {
"n": 112,
"accuracy": 0.8482142857142857,
"deny_recall": 0.8,
"deny_precision": 0.851063829787234,
"false_approve_rate": 0.2,
"false_deny_rate": 0.11290322580645161,
"f1": 0.8247422680412372,
"balanced_accuracy": 0.8435483870967742,
"auroc": 0.9254838709677419
},
"Korean": {
"n": 97,
"accuracy": 0.8969072164948454,
"deny_recall": 0.94,
"deny_precision": 0.8703703703703703,
"false_approve_rate": 0.06,
"false_deny_rate": 0.14893617021276595,
"f1": 0.9038461538461539,
"balanced_accuracy": 0.8955319148936169,
"auroc": 0.9570212765957447
},
"Italian": {
"n": 97,
"accuracy": 0.9381443298969072,
"deny_recall": 0.9787234042553191,
"deny_precision": 0.9019607843137255,
"false_approve_rate": 0.02127659574468085,
"false_deny_rate": 0.1,
"f1": 0.9387755102040817,
"balanced_accuracy": 0.9393617021276596,
"auroc": 0.9889361702127659
},
"French": {
"n": 91,
"accuracy": 0.8571428571428571,
"deny_recall": 0.8666666666666667,
"deny_precision": 0.8478260869565217,
"false_approve_rate": 0.13333333333333333,
"false_deny_rate": 0.15217391304347827,
"f1": 0.8571428571428571,
"balanced_accuracy": 0.8572463768115942,
"auroc": 0.9545893719806764
},
"Russian": {
"n": 90,
"accuracy": 0.8666666666666667,
"deny_recall": 0.8936170212765957,
"deny_precision": 0.8571428571428571,
"false_approve_rate": 0.10638297872340426,
"false_deny_rate": 0.16279069767441862,
"f1": 0.875,
"balanced_accuracy": 0.8654131618010885,
"auroc": 0.9554675903018308
},
"Spanish": {
"n": 89,
"accuracy": 0.8426966292134831,
"deny_recall": 0.8461538461538461,
"deny_precision": 0.8048780487804879,
"false_approve_rate": 0.15384615384615385,
"false_deny_rate": 0.16,
"f1": 0.8250000000000001,
"balanced_accuracy": 0.8430769230769231,
"auroc": 0.9107692307692308
},
"Mandarin Chinese": {
"n": 83,
"accuracy": 0.8433734939759037,
"deny_recall": 0.9166666666666666,
"deny_precision": 0.7674418604651163,
"false_approve_rate": 0.08333333333333333,
"false_deny_rate": 0.2127659574468085,
"f1": 0.8354430379746837,
"balanced_accuracy": 0.8519503546099291,
"auroc": 0.9550827423167848
},
"German": {
"n": 82,
"accuracy": 0.8658536585365854,
"deny_recall": 0.8,
"deny_precision": 0.875,
"false_approve_rate": 0.2,
"false_deny_rate": 0.0851063829787234,
"f1": 0.8358208955223881,
"balanced_accuracy": 0.8574468085106384,
"auroc": 0.9556231003039514
},
"Hindi": {
"n": 78,
"accuracy": 0.8205128205128205,
"deny_recall": 0.8461538461538461,
"deny_precision": 0.8048780487804879,
"false_approve_rate": 0.15384615384615385,
"false_deny_rate": 0.20512820512820512,
"f1": 0.8250000000000001,
"balanced_accuracy": 0.8205128205128205,
"auroc": 0.9105851413543721
},
"Portuguese": {
"n": 77,
"accuracy": 0.8311688311688312,
"deny_recall": 0.7837837837837838,
"deny_precision": 0.8529411764705882,
"false_approve_rate": 0.21621621621621623,
"false_deny_rate": 0.125,
"f1": 0.8169014084507041,
"balanced_accuracy": 0.8293918918918919,
"auroc": 0.9378378378378379
}
},
"baseline_keyword": {
"n": 3000,
"accuracy": 0.5796666666666667,
"deny_recall": 0.4061384725196288,
"deny_precision": 0.5701402805611222,
"false_approve_rate": 0.5938615274803711,
"false_deny_rate": 0.2682926829268293,
"f1": 0.4743643184660275,
"balanced_accuracy": 0.5689228947963998,
"auroc": 0.5689228947963998
},
"latency": {
"batched_mean_s_per_item": 0.061453274965286256,
"batched_p50_s": 0.0031373524906659366,
"single_call": {
"p50_len": {
"tokens": 386,
"ms": 7.232022285461426
},
"p90_len": {
"tokens": 13205,
"ms": 149.2302656173706
}
}
}
}