karlexmarin's picture
feat: ship paper artefacts + CLI diagnostic alongside browser tool
535348a
raw
history blame
22 kB
{
"results": {
"crows_pairs_french_physical_appearance": {
"likelihood_difference": 8.304253472222221,
"likelihood_difference_stderr": 1.1645432805257396,
"pct_stereotype": 0.5694444444444444,
"pct_stereotype_stderr": 0.05876396677084613
},
"logiqa": {
"acc": 0.20276497695852536,
"acc_stderr": 0.015770046635584567,
"acc_norm": 0.22734254992319508,
"acc_norm_stderr": 0.016439067675117765
},
"hendrycksTest-miscellaneous": {
"acc": 0.25287356321839083,
"acc_stderr": 0.015543377313719681,
"acc_norm": 0.26309067688378035,
"acc_norm_stderr": 0.015745497169049043
},
"hendrycksTest-machine_learning": {
"acc": 0.1875,
"acc_stderr": 0.0370468111477387,
"acc_norm": 0.25,
"acc_norm_stderr": 0.04109974682633932
},
"crows_pairs_french_nationality": {
"likelihood_difference": 8.420701581027668,
"likelihood_difference_stderr": 0.40347811300443653,
"pct_stereotype": 0.6047430830039525,
"pct_stereotype_stderr": 0.030798170848773863
},
"hendrycksTest-high_school_statistics": {
"acc": 0.2222222222222222,
"acc_stderr": 0.02835321286686343,
"acc_norm": 0.2638888888888889,
"acc_norm_stderr": 0.03005820270430985
},
"hendrycksTest-college_mathematics": {
"acc": 0.19,
"acc_stderr": 0.03942772444036624,
"acc_norm": 0.24,
"acc_norm_stderr": 0.04292346959909283
},
"hendrycksTest-high_school_geography": {
"acc": 0.22727272727272727,
"acc_stderr": 0.029857515673386417,
"acc_norm": 0.2474747474747475,
"acc_norm_stderr": 0.03074630074212451
},
"hendrycksTest-college_biology": {
"acc": 0.20833333333333334,
"acc_stderr": 0.03396116205845335,
"acc_norm": 0.24305555555555555,
"acc_norm_stderr": 0.03586879280080341
},
"hendrycksTest-philosophy": {
"acc": 0.27009646302250806,
"acc_stderr": 0.025218040373410616,
"acc_norm": 0.27009646302250806,
"acc_norm_stderr": 0.02521804037341062
},
"hendrycksTest-elementary_mathematics": {
"acc": 0.20634920634920634,
"acc_stderr": 0.02084229093011467,
"acc_norm": 0.2222222222222222,
"acc_norm_stderr": 0.021411684393694185
},
"crows_pairs_english_disability": {
"likelihood_difference": 7.021634615384615,
"likelihood_difference_stderr": 0.97111821919133,
"pct_stereotype": 0.47692307692307695,
"pct_stereotype_stderr": 0.062433396464415106
},
"crows_pairs_english_sexual_orientation": {
"likelihood_difference": 5.677083333333333,
"likelihood_difference_stderr": 0.6378864031151475,
"pct_stereotype": 0.5053763440860215,
"pct_stereotype_stderr": 0.05212558986469174
},
"crows_pairs_french_socioeconomic": {
"likelihood_difference": 10.822225765306122,
"likelihood_difference_stderr": 0.6868426421668223,
"pct_stereotype": 0.3010204081632653,
"pct_stereotype_stderr": 0.0328483010552734
},
"hendrycksTest-moral_scenarios": {
"acc": 0.23798882681564246,
"acc_stderr": 0.014242630070574915,
"acc_norm": 0.27262569832402234,
"acc_norm_stderr": 0.014893391735249588
},
"hendrycksTest-security_studies": {
"acc": 0.3183673469387755,
"acc_stderr": 0.02982253379398205,
"acc_norm": 0.22857142857142856,
"acc_norm_stderr": 0.02688214492230775
},
"crows_pairs_english_religion": {
"likelihood_difference": 5.092623873873874,
"likelihood_difference_stderr": 0.5870374592750236,
"pct_stereotype": 0.5045045045045045,
"pct_stereotype_stderr": 0.047671194793956616
},
"hendrycksTest-human_sexuality": {
"acc": 0.32061068702290074,
"acc_stderr": 0.04093329229834278,
"acc_norm": 0.31297709923664124,
"acc_norm_stderr": 0.04066962905677697
},
"crows_pairs_english_autre": {
"likelihood_difference": 5.673295454545454,
"likelihood_difference_stderr": 2.5462810016897417,
"pct_stereotype": 0.45454545454545453,
"pct_stereotype_stderr": 0.15745916432444335
},
"hendrycksTest-nutrition": {
"acc": 0.1895424836601307,
"acc_stderr": 0.022442358263336216,
"acc_norm": 0.2647058823529412,
"acc_norm_stderr": 0.02526169121972948
},
"wsc": {
"acc": 0.6346153846153846,
"acc_stderr": 0.0474473339327792
},
"crows_pairs_french": {
"likelihood_difference": 8.28959824090638,
"likelihood_difference_stderr": 0.1813110538245477,
"pct_stereotype": 0.5724508050089445,
"pct_stereotype_stderr": 0.012084400901134948
},
"hendrycksTest-marketing": {
"acc": 0.20512820512820512,
"acc_stderr": 0.02645350805404033,
"acc_norm": 0.23076923076923078,
"acc_norm_stderr": 0.027601921381417607
},
"hendrycksTest-high_school_government_and_politics": {
"acc": 0.20725388601036268,
"acc_stderr": 0.02925282329180363,
"acc_norm": 0.27461139896373055,
"acc_norm_stderr": 0.03221024508041156
},
"hendrycksTest-professional_law": {
"acc": 0.23533246414602346,
"acc_stderr": 0.010834432543912219,
"acc_norm": 0.2653194263363755,
"acc_norm_stderr": 0.011276198843958866
},
"hendrycksTest-college_chemistry": {
"acc": 0.2,
"acc_stderr": 0.04020151261036846,
"acc_norm": 0.23,
"acc_norm_stderr": 0.042295258468165044
},
"hendrycksTest-us_foreign_policy": {
"acc": 0.23,
"acc_stderr": 0.04229525846816505,
"acc_norm": 0.28,
"acc_norm_stderr": 0.04512608598542129
},
"hendrycksTest-college_medicine": {
"acc": 0.2254335260115607,
"acc_stderr": 0.03186209851641144,
"acc_norm": 0.27167630057803466,
"acc_norm_stderr": 0.03391750322321659
},
"hendrycksTest-high_school_physics": {
"acc": 0.2119205298013245,
"acc_stderr": 0.03336767086567977,
"acc_norm": 0.26490066225165565,
"acc_norm_stderr": 0.03603038545360384
},
"hendrycksTest-high_school_biology": {
"acc": 0.19032258064516128,
"acc_stderr": 0.022331707611823078,
"acc_norm": 0.2161290322580645,
"acc_norm_stderr": 0.023415293433568515
},
"lambada_openai": {
"ppl": 1665636.981895382,
"ppl_stderr": 131507.1881823213,
"acc": 0.0,
"acc_stderr": 0.0
},
"crows_pairs_french_sexual_orientation": {
"likelihood_difference": 13.809065934065934,
"likelihood_difference_stderr": 0.8840201832791948,
"pct_stereotype": 0.8021978021978022,
"pct_stereotype_stderr": 0.04198895203196222
},
"hendrycksTest-abstract_algebra": {
"acc": 0.2,
"acc_stderr": 0.04020151261036845,
"acc_norm": 0.15,
"acc_norm_stderr": 0.03588702812826372
},
"hendrycksTest-business_ethics": {
"acc": 0.31,
"acc_stderr": 0.04648231987117316,
"acc_norm": 0.35,
"acc_norm_stderr": 0.04793724854411019
},
"hendrycksTest-human_aging": {
"acc": 0.2825112107623318,
"acc_stderr": 0.03021683101150877,
"acc_norm": 0.27802690582959644,
"acc_norm_stderr": 0.030069584874494033
},
"hendrycksTest-moral_disputes": {
"acc": 0.21098265895953758,
"acc_stderr": 0.021966309947043117,
"acc_norm": 0.2023121387283237,
"acc_norm_stderr": 0.021628077380196137
},
"crows_pairs_french_religion": {
"likelihood_difference": 9.863315217391305,
"likelihood_difference_stderr": 0.6835325659384403,
"pct_stereotype": 0.6782608695652174,
"pct_stereotype_stderr": 0.04375199868936841
},
"hendrycksTest-sociology": {
"acc": 0.2736318407960199,
"acc_stderr": 0.031524391865554044,
"acc_norm": 0.26865671641791045,
"acc_norm_stderr": 0.03134328358208954
},
"piqa": {
"acc": 0.5261153427638737,
"acc_stderr": 0.011649900854263423,
"acc_norm": 0.5184983677910773,
"acc_norm_stderr": 0.01165783758381816
},
"arc_challenge": {
"acc": 0.20136518771331058,
"acc_stderr": 0.011718927477444262,
"acc_norm": 0.23890784982935154,
"acc_norm_stderr": 0.012461071376316617
},
"hendrycksTest-high_school_microeconomics": {
"acc": 0.18067226890756302,
"acc_stderr": 0.02499196496660076,
"acc_norm": 0.27310924369747897,
"acc_norm_stderr": 0.028942004040998167
},
"arc_easy": {
"acc": 0.2676767676767677,
"acc_stderr": 0.009085000147099353,
"acc_norm": 0.2756734006734007,
"acc_norm_stderr": 0.009169229476542562
},
"winogrande": {
"acc": 0.49171270718232046,
"acc_stderr": 0.014050555322824192
},
"hendrycksTest-astronomy": {
"acc": 0.18421052631578946,
"acc_stderr": 0.031546980450822305,
"acc_norm": 0.29605263157894735,
"acc_norm_stderr": 0.03715062154998905
},
"hendrycksTest-professional_accounting": {
"acc": 0.24468085106382978,
"acc_stderr": 0.02564555362226673,
"acc_norm": 0.25886524822695034,
"acc_norm_stderr": 0.026129572527180848
},
"sciq": {
"acc": 0.219,
"acc_stderr": 0.013084731950262024,
"acc_norm": 0.225,
"acc_norm_stderr": 0.013211720158614751
},
"hendrycksTest-high_school_computer_science": {
"acc": 0.23,
"acc_stderr": 0.042295258468165044,
"acc_norm": 0.29,
"acc_norm_stderr": 0.045604802157206845
},
"hendrycksTest-conceptual_physics": {
"acc": 0.3021276595744681,
"acc_stderr": 0.030017554471880557,
"acc_norm": 0.2723404255319149,
"acc_norm_stderr": 0.029101290698386698
},
"hendrycksTest-medical_genetics": {
"acc": 0.25,
"acc_stderr": 0.04351941398892446,
"acc_norm": 0.25,
"acc_norm_stderr": 0.04351941398892446
},
"hendrycksTest-high_school_european_history": {
"acc": 0.15757575757575756,
"acc_stderr": 0.02845038880528435,
"acc_norm": 0.22424242424242424,
"acc_norm_stderr": 0.03256866661681102
},
"hendrycksTest-management": {
"acc": 0.23300970873786409,
"acc_stderr": 0.04185832598928315,
"acc_norm": 0.2815533980582524,
"acc_norm_stderr": 0.04453254836326466
},
"crows_pairs_english": {
"likelihood_difference": 5.005841905187835,
"likelihood_difference_stderr": 0.1594255961602287,
"pct_stereotype": 0.4442456768038163,
"pct_stereotype_stderr": 0.012137130534698495
},
"hendrycksTest-public_relations": {
"acc": 0.2727272727272727,
"acc_stderr": 0.04265792110940588,
"acc_norm": 0.16363636363636364,
"acc_norm_stderr": 0.03543433054298678
},
"hendrycksTest-computer_security": {
"acc": 0.24,
"acc_stderr": 0.042923469599092816,
"acc_norm": 0.29,
"acc_norm_stderr": 0.045604802157206845
},
"hendrycksTest-virology": {
"acc": 0.18674698795180722,
"acc_stderr": 0.030338749144500618,
"acc_norm": 0.23493975903614459,
"acc_norm_stderr": 0.03300533186128922
},
"hendrycksTest-high_school_mathematics": {
"acc": 0.14814814814814814,
"acc_stderr": 0.02165977842211803,
"acc_norm": 0.22592592592592592,
"acc_norm_stderr": 0.02549753263960955
},
"hendrycksTest-anatomy": {
"acc": 0.18518518518518517,
"acc_stderr": 0.03355677216313142,
"acc_norm": 0.26666666666666666,
"acc_norm_stderr": 0.038201699145179055
},
"hendrycksTest-formal_logic": {
"acc": 0.2857142857142857,
"acc_stderr": 0.04040610178208841,
"acc_norm": 0.2777777777777778,
"acc_norm_stderr": 0.040061680838488774
},
"hendrycksTest-jurisprudence": {
"acc": 0.1574074074074074,
"acc_stderr": 0.035207039905179656,
"acc_norm": 0.23148148148148148,
"acc_norm_stderr": 0.04077494709252627
},
"hendrycksTest-international_law": {
"acc": 0.09090909090909091,
"acc_stderr": 0.026243194054073878,
"acc_norm": 0.2727272727272727,
"acc_norm_stderr": 0.04065578140908705
},
"crows_pairs_french_autre": {
"likelihood_difference": 6.225961538461538,
"likelihood_difference_stderr": 1.0845778061093316,
"pct_stereotype": 0.6153846153846154,
"pct_stereotype_stderr": 0.14044168141158106
},
"hendrycksTest-global_facts": {
"acc": 0.23,
"acc_stderr": 0.04229525846816506,
"acc_norm": 0.23,
"acc_norm_stderr": 0.04229525846816506
},
"crows_pairs_english_gender": {
"likelihood_difference": 4.482080078125,
"likelihood_difference_stderr": 0.4507987456254625,
"pct_stereotype": 0.54375,
"pct_stereotype_stderr": 0.027887252708654657
},
"crows_pairs_english_nationality": {
"likelihood_difference": 6.005063657407407,
"likelihood_difference_stderr": 0.4386043275416813,
"pct_stereotype": 0.3194444444444444,
"pct_stereotype_stderr": 0.0317987634217685
},
"crows_pairs_english_race_color": {
"likelihood_difference": 4.779035433070866,
"likelihood_difference_stderr": 0.26206678827654106,
"pct_stereotype": 0.3346456692913386,
"pct_stereotype_stderr": 0.02095632470166831
},
"crows_pairs_french_race_color": {
"likelihood_difference": 7.161277173913043,
"likelihood_difference_stderr": 0.26238666011376527,
"pct_stereotype": 0.7108695652173913,
"pct_stereotype_stderr": 0.02116096760624947
},
"crows_pairs_french_disability": {
"likelihood_difference": 12.47064393939394,
"likelihood_difference_stderr": 1.234102889806993,
"pct_stereotype": 0.3333333333333333,
"pct_stereotype_stderr": 0.0584705346204686
},
"hendrycksTest-high_school_world_history": {
"acc": 0.19831223628691982,
"acc_stderr": 0.025955020841621115,
"acc_norm": 0.2616033755274262,
"acc_norm_stderr": 0.028609516716994934
},
"hendrycksTest-logical_fallacies": {
"acc": 0.20245398773006135,
"acc_stderr": 0.031570650789119,
"acc_norm": 0.3006134969325153,
"acc_norm_stderr": 0.03602511318806771
},
"hendrycksTest-high_school_chemistry": {
"acc": 0.18719211822660098,
"acc_stderr": 0.027444924966882618,
"acc_norm": 0.21674876847290642,
"acc_norm_stderr": 0.02899033125251624
},
"hendrycksTest-college_computer_science": {
"acc": 0.23,
"acc_stderr": 0.04229525846816506,
"acc_norm": 0.21,
"acc_norm_stderr": 0.040936018074033256
},
"crows_pairs_french_age": {
"likelihood_difference": 4.382638888888889,
"likelihood_difference_stderr": 0.563720955887399,
"pct_stereotype": 0.4666666666666667,
"pct_stereotype_stderr": 0.05288198530254015
},
"hendrycksTest-high_school_macroeconomics": {
"acc": 0.21025641025641026,
"acc_stderr": 0.020660597485026938,
"acc_norm": 0.26153846153846155,
"acc_norm_stderr": 0.022282141204204426
},
"hendrycksTest-professional_psychology": {
"acc": 0.24183006535947713,
"acc_stderr": 0.017322789207784326,
"acc_norm": 0.26633986928104575,
"acc_norm_stderr": 0.017883188134667164
},
"hendrycksTest-econometrics": {
"acc": 0.21929824561403508,
"acc_stderr": 0.03892431106518752,
"acc_norm": 0.2719298245614035,
"acc_norm_stderr": 0.04185774424022057
},
"crows_pairs_english_age": {
"likelihood_difference": 3.299793956043956,
"likelihood_difference_stderr": 0.48606624275634924,
"pct_stereotype": 0.4835164835164835,
"pct_stereotype_stderr": 0.05267597952306975
},
"hendrycksTest-clinical_knowledge": {
"acc": 0.18490566037735848,
"acc_stderr": 0.02389335183446432,
"acc_norm": 0.30943396226415093,
"acc_norm_stderr": 0.028450154794118627
},
"hendrycksTest-college_physics": {
"acc": 0.17647058823529413,
"acc_stderr": 0.0379328118530781,
"acc_norm": 0.17647058823529413,
"acc_norm_stderr": 0.037932811853078105
},
"crows_pairs_english_physical_appearance": {
"likelihood_difference": 5.136284722222222,
"likelihood_difference_stderr": 0.6655017702667131,
"pct_stereotype": 0.5277777777777778,
"pct_stereotype_stderr": 0.05924743948371487
},
"crows_pairs_french_gender": {
"likelihood_difference": 6.4443146417445485,
"likelihood_difference_stderr": 0.3439159663666914,
"pct_stereotype": 0.48909657320872274,
"pct_stereotype_stderr": 0.02794420307081864
},
"hendrycksTest-high_school_us_history": {
"acc": 0.19117647058823528,
"acc_stderr": 0.02759917430064077,
"acc_norm": 0.23039215686274508,
"acc_norm_stderr": 0.029554292605695063
},
"hendrycksTest-electrical_engineering": {
"acc": 0.2413793103448276,
"acc_stderr": 0.03565998174135302,
"acc_norm": 0.2620689655172414,
"acc_norm_stderr": 0.036646663372252565
},
"crows_pairs_english_socioeconomic": {
"likelihood_difference": 5.01858552631579,
"likelihood_difference_stderr": 0.4299366436044498,
"pct_stereotype": 0.5842105263157895,
"pct_stereotype_stderr": 0.0358501132552001
},
"hendrycksTest-prehistory": {
"acc": 0.2623456790123457,
"acc_stderr": 0.024477222856135118,
"acc_norm": 0.24691358024691357,
"acc_norm_stderr": 0.023993501709042124
},
"hendrycksTest-high_school_psychology": {
"acc": 0.23669724770642203,
"acc_stderr": 0.01822407811729907,
"acc_norm": 0.24770642201834864,
"acc_norm_stderr": 0.018508143602547815
},
"hendrycksTest-professional_medicine": {
"acc": 0.26838235294117646,
"acc_stderr": 0.026917481224377232,
"acc_norm": 0.28308823529411764,
"acc_norm_stderr": 0.027365861131513805
},
"hendrycksTest-world_religions": {
"acc": 0.1871345029239766,
"acc_stderr": 0.029913127232368043,
"acc_norm": 0.23976608187134502,
"acc_norm_stderr": 0.032744852119469564
}
},
"versions": {
"crows_pairs_french_physical_appearance": 0,
"logiqa": 0,
"hendrycksTest-miscellaneous": 0,
"hendrycksTest-machine_learning": 0,
"crows_pairs_french_nationality": 0,
"hendrycksTest-high_school_statistics": 0,
"hendrycksTest-college_mathematics": 0,
"hendrycksTest-high_school_geography": 0,
"hendrycksTest-college_biology": 0,
"hendrycksTest-philosophy": 0,
"hendrycksTest-elementary_mathematics": 0,
"crows_pairs_english_disability": 0,
"crows_pairs_english_sexual_orientation": 0,
"crows_pairs_french_socioeconomic": 0,
"hendrycksTest-moral_scenarios": 0,
"hendrycksTest-security_studies": 0,
"crows_pairs_english_religion": 0,
"hendrycksTest-human_sexuality": 0,
"crows_pairs_english_autre": 0,
"hendrycksTest-nutrition": 0,
"wsc": 0,
"crows_pairs_french": 0,
"hendrycksTest-marketing": 0,
"hendrycksTest-high_school_government_and_politics": 0,
"hendrycksTest-professional_law": 0,
"hendrycksTest-college_chemistry": 0,
"hendrycksTest-us_foreign_policy": 0,
"hendrycksTest-college_medicine": 0,
"hendrycksTest-high_school_physics": 0,
"hendrycksTest-high_school_biology": 0,
"lambada_openai": 0,
"crows_pairs_french_sexual_orientation": 0,
"hendrycksTest-abstract_algebra": 0,
"hendrycksTest-business_ethics": 0,
"hendrycksTest-human_aging": 0,
"hendrycksTest-moral_disputes": 0,
"crows_pairs_french_religion": 0,
"hendrycksTest-sociology": 0,
"piqa": 0,
"arc_challenge": 0,
"hendrycksTest-high_school_microeconomics": 0,
"arc_easy": 0,
"winogrande": 0,
"hendrycksTest-astronomy": 0,
"hendrycksTest-professional_accounting": 0,
"sciq": 0,
"hendrycksTest-high_school_computer_science": 0,
"hendrycksTest-conceptual_physics": 0,
"hendrycksTest-medical_genetics": 0,
"hendrycksTest-high_school_european_history": 0,
"hendrycksTest-management": 0,
"crows_pairs_english": 0,
"hendrycksTest-public_relations": 0,
"hendrycksTest-computer_security": 0,
"hendrycksTest-virology": 0,
"hendrycksTest-high_school_mathematics": 0,
"hendrycksTest-anatomy": 0,
"hendrycksTest-formal_logic": 0,
"hendrycksTest-jurisprudence": 0,
"hendrycksTest-international_law": 0,
"crows_pairs_french_autre": 0,
"hendrycksTest-global_facts": 0,
"crows_pairs_english_gender": 0,
"crows_pairs_english_nationality": 0,
"crows_pairs_english_race_color": 0,
"crows_pairs_french_race_color": 0,
"crows_pairs_french_disability": 0,
"hendrycksTest-high_school_world_history": 0,
"hendrycksTest-logical_fallacies": 0,
"hendrycksTest-high_school_chemistry": 0,
"hendrycksTest-college_computer_science": 0,
"crows_pairs_french_age": 0,
"hendrycksTest-high_school_macroeconomics": 0,
"hendrycksTest-professional_psychology": 0,
"hendrycksTest-econometrics": 0,
"crows_pairs_english_age": 0,
"hendrycksTest-clinical_knowledge": 0,
"hendrycksTest-college_physics": 0,
"crows_pairs_english_physical_appearance": 0,
"crows_pairs_french_gender": 0,
"hendrycksTest-high_school_us_history": 0,
"hendrycksTest-electrical_engineering": 0,
"crows_pairs_english_socioeconomic": 0,
"hendrycksTest-prehistory": 0,
"hendrycksTest-high_school_psychology": 0,
"hendrycksTest-professional_medicine": 0,
"hendrycksTest-world_religions": 0
},
"config": {
"model": "hf-causal",
"model_args": "use_accelerate=True,pretrained=EleutherAI/pythia-v1.1-70m,revision=step128",
"num_fewshot": 0,
"batch_size": 32,
"device": null,
"no_cache": true,
"limit": null,
"bootstrap_iters": 100000,
"description_dict": {}
}
}