multilingual-toxicity-base / eval /baselines.json
Horizon-Labs's picture
Baselines: + gravitee distilbert, xlmr-large v2
2927db8 verified
Raw History Blame Contribute Delete
15.8 kB
{
"unitary/unbiased-toxic-roberta": {
"labels": [
"toxicity",
"severe_toxicity",
"obscene",
"identity_attack",
"insult",
"threat",
"sexual_explicit",
"male",
"female",
"homosexual_gay_or_lesbian",
"christian",
"jewish",
"muslim",
"black",
"white",
"psychiatric_or_mental_illness"
],
"mapped": [
"toxicity",
"severe_toxicity",
"obscene",
"identity_attack",
"insult",
"threat",
"sexual_explicit"
],
"civil_test": {
"n": 20000,
"auc": {
"toxicity": 0.9690025757582605,
"obscene": 0.9909355527122459,
"threat": 0.9848926893092385,
"insult": 0.9799812478376196,
"identity_attack": 0.986828551415347,
"sexual_explicit": 0.9959576840892516
},
"mean_auc": 0.9845997168536605,
"toxicity_auc": 0.9690025757582605
},
"tdx_am": {
"n": 1000,
"auc": 0.516552,
"f1_05": 0.003992015968063872
},
"tdx_ar": {
"n": 1000,
"auc": 0.466138,
"f1_05": 0.0
},
"tdx_de": {
"n": 1000,
"auc": 0.626188,
"f1_05": 0.03137254901960784
},
"tdx_en": {
"n": 1000,
"auc": 0.996856,
"f1_05": 0.9684638860630722
},
"tdx_es": {
"n": 1000,
"auc": 0.645716,
"f1_05": 0.12962962962962962
},
"tdx_fr": {
"n": 1000,
"auc": 0.728348,
"f1_05": 0.07279693486590039
},
"tdx_he": {
"n": 1000,
"auc": 0.53873,
"f1_05": 0.0
},
"tdx_hi": {
"n": 1000,
"auc": 0.44654,
"f1_05": 0.0
},
"tdx_hin": {
"n": 1000,
"auc": 0.60834,
"f1_05": 0.019762845849802372
},
"tdx_it": {
"n": 1000,
"auc": 0.632824,
"f1_05": 0.06500956022944551
},
"tdx_ja": {
"n": 1000,
"auc": 0.517142,
"f1_05": 0.0
},
"tdx_ru": {
"n": 1000,
"auc": 0.509322,
"f1_05": 0.0
},
"tdx_tt": {
"n": 1000,
"auc": 0.493712,
"f1_05": 0.0
},
"tdx_uk": {
"n": 1000,
"auc": 0.53974,
"f1_05": 0.0
},
"tdx_zh": {
"n": 1000,
"auc": 0.560074,
"f1_05": 0.0
},
"_civil_mean_auc": 0.9845997168536605,
"_civil_toxicity_auc": 0.9690025757582605,
"_tdx_mean_auc": 0.5884148000000001,
"_tdx_mean_f1": 0.08606849477503481
},
"unitary/multilingual-toxic-xlm-roberta": {
"labels": [
"toxic"
],
"mapped": [
"toxicity"
],
"civil_test": {
"n": 20000,
"auc": {
"toxicity": 0.9606788196819926
},
"mean_auc": 0.9606788196819926,
"toxicity_auc": 0.9606788196819926
},
"tdx_am": {
"n": 1000,
"auc": 0.547776,
"f1_05": 0.10657193605683836
},
"tdx_ar": {
"n": 1000,
"auc": 0.412792,
"f1_05": 0.0
},
"tdx_de": {
"n": 1000,
"auc": 0.66235,
"f1_05": 0.16333938294010888
},
"tdx_en": {
"n": 1000,
"auc": 0.99174,
"f1_05": 0.9557344064386318
},
"tdx_es": {
"n": 1000,
"auc": 0.819208,
"f1_05": 0.6459627329192547
},
"tdx_fr": {
"n": 1000,
"auc": 0.98129,
"f1_05": 0.9088960342979635
},
"tdx_he": {
"n": 1000,
"auc": 0.501396,
"f1_05": 0.003976143141153081
},
"tdx_hi": {
"n": 1000,
"auc": 0.644394,
"f1_05": 0.00796812749003984
},
"tdx_hin": {
"n": 1000,
"auc": 0.630578,
"f1_05": 0.18867924528301888
},
"tdx_it": {
"n": 1000,
"auc": 0.818168,
"f1_05": 0.7285429141716567
},
"tdx_ja": {
"n": 1000,
"auc": 0.50647,
"f1_05": 0.0
},
"tdx_ru": {
"n": 1000,
"auc": 0.806734,
"f1_05": 0.6200527704485488
},
"tdx_tt": {
"n": 1000,
"auc": 0.62617,
"f1_05": 0.47579529737206083
},
"tdx_uk": {
"n": 1000,
"auc": 0.679846,
"f1_05": 0.45573294629898403
},
"tdx_zh": {
"n": 1000,
"auc": 0.54716,
"f1_05": 0.015810276679841896
},
"_civil_mean_auc": 0.9606788196819926,
"_civil_toxicity_auc": 0.9606788196819926,
"_tdx_mean_auc": 0.6784047999999999,
"_tdx_mean_f1": 0.35180414756920675
},
"unitary/toxic-bert": {
"labels": [
"toxic",
"severe_toxic",
"obscene",
"threat",
"insult",
"identity_hate"
],
"mapped": [
"toxicity",
"severe_toxicity",
"obscene",
"threat",
"insult",
"identity_attack"
],
"civil_test": {
"n": 20000,
"auc": {
"toxicity": 0.9256542735522112,
"obscene": 0.9686241575896402,
"threat": 0.9446838368037228,
"insult": 0.9367789182331911,
"identity_attack": 0.9550349685566486
},
"mean_auc": 0.9461552309470829,
"toxicity_auc": 0.9256542735522112
},
"tdx_am": {
"n": 1000,
"auc": 0.57372,
"f1_05": 0.023529411764705882
},
"tdx_ar": {
"n": 1000,
"auc": 0.570894,
"f1_05": 0.0
},
"tdx_de": {
"n": 1000,
"auc": 0.648354,
"f1_05": 0.3783783783783784
},
"tdx_en": {
"n": 1000,
"auc": 0.985168,
"f1_05": 0.9402390438247012
},
"tdx_es": {
"n": 1000,
"auc": 0.697046,
"f1_05": 0.17375886524822695
},
"tdx_fr": {
"n": 1000,
"auc": 0.765274,
"f1_05": 0.09433962264150944
},
"tdx_he": {
"n": 1000,
"auc": 0.514606,
"f1_05": 0.0
},
"tdx_hi": {
"n": 1000,
"auc": 0.634886,
"f1_05": 0.003992015968063872
},
"tdx_hin": {
"n": 1000,
"auc": 0.660786,
"f1_05": 0.2598684210526316
},
"tdx_it": {
"n": 1000,
"auc": 0.659214,
"f1_05": 0.2578512396694215
},
"tdx_ja": {
"n": 1000,
"auc": 0.628196,
"f1_05": 0.0
},
"tdx_ru": {
"n": 1000,
"auc": 0.53582,
"f1_05": 0.0
},
"tdx_tt": {
"n": 1000,
"auc": 0.471088,
"f1_05": 0.0
},
"tdx_uk": {
"n": 1000,
"auc": 0.521712,
"f1_05": 0.0
},
"tdx_zh": {
"n": 1000,
"auc": 0.481674,
"f1_05": 0.003992015968063872
},
"_civil_mean_auc": 0.9461552309470829,
"_civil_toxicity_auc": 0.9256542735522112,
"_tdx_mean_auc": 0.6232292,
"_tdx_mean_f1": 0.14239660096771353
},
"textdetox/xlmr-large-toxicity-classifier": {
"labels": [
"neutral",
"toxic"
],
"mapped": [
"toxicity"
],
"civil_test": {
"n": 20000,
"auc": {
"toxicity": 0.8690952803195003
},
"mean_auc": 0.8690952803195003,
"toxicity_auc": 0.8690952803195003
},
"tdx_am": {
"n": 1000,
"auc": 0.981514,
"f1_05": 0.9400196656833825
},
"tdx_ar": {
"n": 1000,
"auc": 0.98943,
"f1_05": 0.9558232931726908
},
"tdx_de": {
"n": 1000,
"auc": 0.996744,
"f1_05": 0.9859154929577465
},
"tdx_en": {
"n": 1000,
"auc": 0.999874,
"f1_05": 0.998998998998999
},
"tdx_es": {
"n": 1000,
"auc": 0.99761,
"f1_05": 0.986986986986987
},
"tdx_fr": {
"n": 1000,
"auc": 0.97144,
"f1_05": 0.9019189765458422
},
"tdx_he": {
"n": 1000,
"auc": 0.701154,
"f1_05": 0.61742006615215
},
"tdx_hi": {
"n": 1000,
"auc": 0.995166,
"f1_05": 0.982982982982983
},
"tdx_hin": {
"n": 1000,
"auc": 0.755944,
"f1_05": 0.7315834038950042
},
"tdx_it": {
"n": 1000,
"auc": 0.80468,
"f1_05": 0.7455731593662628
},
"tdx_ja": {
"n": 1000,
"auc": 0.851862,
"f1_05": 0.6749688667496887
},
"tdx_ru": {
"n": 1000,
"auc": 0.999984,
"f1_05": 0.9960079840319361
},
"tdx_tt": {
"n": 1000,
"auc": 0.80623,
"f1_05": 0.7157360406091371
},
"tdx_uk": {
"n": 1000,
"auc": 0.996054,
"f1_05": 0.988988988988989
},
"tdx_zh": {
"n": 1000,
"auc": 0.979184,
"f1_05": 0.9409409409409409
},
"_civil_mean_auc": 0.8690952803195003,
"_civil_toxicity_auc": 0.8690952803195003,
"_tdx_mean_auc": 0.9217913333333333,
"_tdx_mean_f1": 0.8775910565375159
},
"s-nlp/roberta_toxicity_classifier": {
"labels": [
"neutral",
"toxic"
],
"mapped": [
"toxicity"
],
"civil_test": {
"n": 20000,
"auc": {
"toxicity": 0.9752315080183933
},
"mean_auc": 0.9752315080183933,
"toxicity_auc": 0.9752315080183933
},
"tdx_am": {
"n": 1000,
"auc": 0.573324,
"f1_05": 0.003992015968063872
},
"tdx_ar": {
"n": 1000,
"auc": 0.531116,
"f1_05": 0.0
},
"tdx_de": {
"n": 1000,
"auc": 0.697008,
"f1_05": 0.03536345776031434
},
"tdx_en": {
"n": 1000,
"auc": 0.999946,
"f1_05": 0.999000999000999
},
"tdx_es": {
"n": 1000,
"auc": 0.74753,
"f1_05": 0.16363636363636364
},
"tdx_fr": {
"n": 1000,
"auc": 0.816256,
"f1_05": 0.05825242718446602
},
"tdx_he": {
"n": 1000,
"auc": 0.47915,
"f1_05": 0.0
},
"tdx_hi": {
"n": 1000,
"auc": 0.706628,
"f1_05": 0.0
},
"tdx_hin": {
"n": 1000,
"auc": 0.642148,
"f1_05": 0.023715415019762844
},
"tdx_it": {
"n": 1000,
"auc": 0.692894,
"f1_05": 0.1016949152542373
},
"tdx_ja": {
"n": 1000,
"auc": 0.559616,
"f1_05": 0.0
},
"tdx_ru": {
"n": 1000,
"auc": 0.571404,
"f1_05": 0.0
},
"tdx_tt": {
"n": 1000,
"auc": 0.53281,
"f1_05": 0.0
},
"tdx_uk": {
"n": 1000,
"auc": 0.504392,
"f1_05": 0.0
},
"tdx_zh": {
"n": 1000,
"auc": 0.570522,
"f1_05": 0.0
},
"_civil_mean_auc": 0.9752315080183933,
"_civil_toxicity_auc": 0.9752315080183933,
"_tdx_mean_auc": 0.6416495999999998,
"_tdx_mean_f1": 0.09237703958828046
},
"martin-ha/toxic-comment-model": {
"labels": [
"non-toxic",
"toxic"
],
"mapped": [
"toxicity"
],
"civil_test": {
"n": 20000,
"auc": {
"toxicity": 0.9433817010496016
},
"mean_auc": 0.9433817010496016,
"toxicity_auc": 0.9433817010496016
},
"tdx_am": {
"n": 1000,
"auc": 0.547124,
"f1_05": 0.003992015968063872
},
"tdx_ar": {
"n": 1000,
"auc": 0.46348,
"f1_05": 0.0
},
"tdx_de": {
"n": 1000,
"auc": 0.602732,
"f1_05": 0.24242424242424243
},
"tdx_en": {
"n": 1000,
"auc": 0.932474,
"f1_05": 0.8483516483516483
},
"tdx_es": {
"n": 1000,
"auc": 0.626534,
"f1_05": 0.21453287197231835
},
"tdx_fr": {
"n": 1000,
"auc": 0.723076,
"f1_05": 0.1261595547309833
},
"tdx_he": {
"n": 1000,
"auc": 0.520774,
"f1_05": 0.3836317135549872
},
"tdx_hi": {
"n": 1000,
"auc": 0.504692,
"f1_05": 0.23823529411764705
},
"tdx_hin": {
"n": 1000,
"auc": 0.49692,
"f1_05": 0.12521739130434784
},
"tdx_it": {
"n": 1000,
"auc": 0.609736,
"f1_05": 0.21180555555555555
},
"tdx_ja": {
"n": 1000,
"auc": 0.456664,
"f1_05": 0.0
},
"tdx_ru": {
"n": 1000,
"auc": 0.57655,
"f1_05": 0.003992015968063872
},
"tdx_tt": {
"n": 1000,
"auc": 0.46656,
"f1_05": 0.0
},
"tdx_uk": {
"n": 1000,
"auc": 0.541306,
"f1_05": 0.00398406374501992
},
"tdx_zh": {
"n": 1000,
"auc": 0.518432,
"f1_05": 0.0
},
"_civil_mean_auc": 0.9433817010496016,
"_civil_toxicity_auc": 0.9433817010496016,
"_tdx_mean_auc": 0.5724702666666668,
"_tdx_mean_f1": 0.16015509117952517
},
"citizenlab/distilbert-base-multilingual-cased-toxicity": {
"labels": [
"toxic",
"not_toxic"
],
"mapped": [
"toxicity"
],
"civil_test": {
"n": 20000,
"auc": {
"toxicity": 0.8433979621843186
},
"mean_auc": 0.8433979621843186,
"toxicity_auc": 0.8433979621843186
},
"tdx_am": {
"n": 1000,
"auc": 0.562258,
"f1_05": 0.011904761904761904
},
"tdx_ar": {
"n": 1000,
"auc": 0.557862,
"f1_05": 0.19795221843003413
},
"tdx_de": {
"n": 1000,
"auc": 0.726776,
"f1_05": 0.4568245125348189
},
"tdx_en": {
"n": 1000,
"auc": 0.932494,
"f1_05": 0.825503355704698
},
"tdx_es": {
"n": 1000,
"auc": 0.790694,
"f1_05": 0.353870458135861
},
"tdx_fr": {
"n": 1000,
"auc": 0.91149,
"f1_05": 0.4737631184407796
},
"tdx_he": {
"n": 1000,
"auc": 0.593274,
"f1_05": 0.06060606060606061
},
"tdx_hi": {
"n": 1000,
"auc": 0.78338,
"f1_05": 0.031496062992125984
},
"tdx_hin": {
"n": 1000,
"auc": 0.659062,
"f1_05": 0.49624060150375937
},
"tdx_it": {
"n": 1000,
"auc": 0.730984,
"f1_05": 0.5940828402366863
},
"tdx_ja": {
"n": 1000,
"auc": 0.65038,
"f1_05": 0.07590132827324478
},
"tdx_ru": {
"n": 1000,
"auc": 0.729002,
"f1_05": 0.292358803986711
},
"tdx_tt": {
"n": 1000,
"auc": 0.550798,
"f1_05": 0.06691449814126393
},
"tdx_uk": {
"n": 1000,
"auc": 0.654464,
"f1_05": 0.1170018281535649
},
"tdx_zh": {
"n": 1000,
"auc": 0.642762,
"f1_05": 0.1038961038961039
},
"_civil_mean_auc": 0.8433979621843186,
"_civil_toxicity_auc": 0.8433979621843186,
"_tdx_mean_auc": 0.6983786666666668,
"_tdx_mean_f1": 0.2772211035293649
},
"gravitee-io/distilbert-multilingual-toxicity-classifier": {
"labels": [
"not-toxic",
"toxic"
],
"mapped": [
"toxicity"
],
"civil_test": {
"n": 20000,
"auc": {
"toxicity": 0.8089855626873078
},
"mean_auc": 0.8089855626873078,
"toxicity_auc": 0.8089855626873078
},
"tdx_am": {
"n": 1000,
"auc": 0.647036,
"f1_05": 0.684044233807267
},
"tdx_ar": {
"n": 1000,
"auc": 0.91521,
"f1_05": 0.8431372549019608
},
"tdx_de": {
"n": 1000,
"auc": 0.981946,
"f1_05": 0.9219562955254943
},
"tdx_en": {
"n": 1000,
"auc": 0.997914,
"f1_05": 0.9830508474576272
},
"tdx_es": {
"n": 1000,
"auc": 0.981574,
"f1_05": 0.9387755102040817
},
"tdx_fr": {
"n": 1000,
"auc": 0.99783,
"f1_05": 0.9810568295114656
},
"tdx_he": {
"n": 1000,
"auc": 0.913612,
"f1_05": 0.8340425531914893
},
"tdx_hi": {
"n": 1000,
"auc": 0.988272,
"f1_05": 0.955193482688391
},
"tdx_hin": {
"n": 1000,
"auc": 0.902658,
"f1_05": 0.8339622641509434
},
"tdx_it": {
"n": 1000,
"auc": 0.958374,
"f1_05": 0.8951841359773371
},
"tdx_ja": {
"n": 1000,
"auc": 0.975844,
"f1_05": 0.9194915254237288
},
"tdx_ru": {
"n": 1000,
"auc": 0.99792,
"f1_05": 0.9830845771144279
},
"tdx_tt": {
"n": 1000,
"auc": 0.991056,
"f1_05": 0.9592850049652433
},
"tdx_uk": {
"n": 1000,
"auc": 0.986376,
"f1_05": 0.9416498993963782
},
"tdx_zh": {
"n": 1000,
"auc": 0.87896,
"f1_05": 0.7629310344827587
},
"_civil_mean_auc": 0.8089855626873078,
"_civil_toxicity_auc": 0.8089855626873078,
"_tdx_mean_auc": 0.9409721333333334,
"_tdx_mean_f1": 0.8957896965865728
},
"textdetox/xlmr-large-toxicity-classifier-v2": {
"labels": [
"label_0",
"label_1"
],
"mapped": [
"toxicity"
],
"civil_test": {
"n": 20000,
"auc": {
"toxicity": 0.8947823432223221
},
"mean_auc": 0.8947823432223221,
"toxicity_auc": 0.8947823432223221
},
"tdx_am": {
"n": 1000,
"auc": 0.975296,
"f1_05": 0.9230769230769231
},
"tdx_ar": {
"n": 1000,
"auc": 0.98156,
"f1_05": 0.9408224674022067
},
"tdx_de": {
"n": 1000,
"auc": 0.994426,
"f1_05": 0.981018981018981
},
"tdx_en": {
"n": 1000,
"auc": 0.999892,
"f1_05": 0.9959919839679359
},
"tdx_es": {
"n": 1000,
"auc": 0.999264,
"f1_05": 0.9890329012961117
},
"tdx_fr": {
"n": 1000,
"auc": 0.99936,
"f1_05": 0.998003992015968
},
"tdx_he": {
"n": 1000,
"auc": 0.96858,
"f1_05": 0.9126016260162602
},
"tdx_hi": {
"n": 1000,
"auc": 0.993964,
"f1_05": 0.9779559118236473
},
"tdx_hin": {
"n": 1000,
"auc": 0.965486,
"f1_05": 0.9177403369672944
},
"tdx_it": {
"n": 1000,
"auc": 0.990784,
"f1_05": 0.9690927218344965
},
"tdx_ja": {
"n": 1000,
"auc": 0.984078,
"f1_05": 0.9562563580874873
},
"tdx_ru": {
"n": 1000,
"auc": 0.999964,
"f1_05": 0.996
},
"tdx_tt": {
"n": 1000,
"auc": 0.994964,
"f1_05": 0.9726443768996961
},
"tdx_uk": {
"n": 1000,
"auc": 0.996866,
"f1_05": 0.984984984984985
},
"tdx_zh": {
"n": 1000,
"auc": 0.958476,
"f1_05": 0.8995024875621891
},
"_civil_mean_auc": 0.8947823432223221,
"_civil_toxicity_auc": 0.8947823432223221,
"_tdx_mean_auc": 0.986864,
"_tdx_mean_f1": 0.9609817368636122
}
}