{"id":"W3132011002","doi":"10.48550/arxiv.2105.05541","title":"Evaluating Gender Bias in Natural Language Inference","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Debiasing; Gender bias; Task (project management); Premise; Inference; Computer science; Natural (archaeology); Natural language processing; Artificial intelligence; Psychology; Social psychology; Cognitive psychology; Linguistics; Geography","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.04224915,0.001440314,0.0009785873,0.002272511,0.001160007,0.002961632,0.002520952,0.002536907,0.003593647],"category_scores_gemma":[0.1332451,0.0005247531,0.0009027683,0.001380683,0.001919854,0.006701241,0.003328563,0.003547085,0.001861018],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001882541,"about_ca_system_score_gemma":0.001870326,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005521363,"about_ca_topic_score_gemma":0.008333063,"domain_scores_codex":[0.9749615,0.01582684,0.001183638,0.003388865,0.004018818,0.0006203245],"domain_scores_gemma":[0.829726,0.1474099,0.004641854,0.01026045,0.006727359,0.001234389],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.004315295,0.001398495,0.1855644,0.002032031,0.001210176,0.000720196,0.004537342,0.1267524,0.02262329,0.02695777,0.03456929,0.5893193],"study_design_scores_gemma":[0.0002185388,0.0009516386,0.03421635,0.0004393512,0.0002654164,0.0006345225,0.002578405,0.8248323,0.0429447,0.07644659,0.01630259,0.0001695513],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6894932,0.005444693,0.2694376,0.005562654,0.0008757749,0.0005730543,0.005150615,0.005559079,0.01790333],"genre_scores_gemma":[0.905234,0.0005091661,0.08352478,0.001086505,0.0001913715,0.000196788,0.00672553,0.0003847257,0.002147124],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.04224915,"threshold_uncertainty_score":0.2234375,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1913847993683349,"score_gpt":0.2933840641489097,"score_spread":0.1019992647805749,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}