{"id":"W2952349219","doi":"10.18653/v1/p19-1166","title":"Understanding Undesirable Word Embedding Associations","year":2019,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":121,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; University of Toronto","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Word (group theory); Debiasing; Association (psychology); Subspace topology; Computer science; Embedding; Projection (relational algebra); Natural language processing; Product (mathematics); Artificial intelligence; Association test; Word Association; Word embedding; Matrix decomposition; Speech recognition; Algorithm; Mathematics; Psychology; Social psychology","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002338613,0.00005765333,0.00008046979,0.0000643696,0.00009300069,0.0001691673,0.0003513232,0.00003641106,0.0001141348],"category_scores_gemma":[0.00001925232,0.0000553113,0.00003274164,0.000214227,0.000004004334,0.0004716824,0.0001206431,0.00007613143,0.0002717236],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002489279,"about_ca_system_score_gemma":0.00004083725,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000227306,"about_ca_topic_score_gemma":0.000009757421,"domain_scores_codex":[0.999268,0.00002046799,0.0001172339,0.0002138233,0.0001766748,0.0002037763],"domain_scores_gemma":[0.9994757,0.0001011771,0.00004005724,0.0003246694,0.00001890209,0.00003949289],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[1.621437e-7,0.000006668577,0.003793439,0.000002456995,0.000007459438,6.311124e-7,0.0002228589,0.005246966,0.0002104037,0.9894161,0.0004126609,0.0006801438],"study_design_scores_gemma":[0.0001711448,0.000008899353,0.0005418449,0.00001123365,0.000001921285,0.000001434975,0.0002016964,0.9229667,0.0001780887,0.07528222,0.0005039925,0.0001308908],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.00826295,0.000007741762,0.8918086,0.0006840093,0.0003509633,0.00006977641,1.981981e-7,0.0002022069,0.09861358],"genre_scores_gemma":[0.9074612,0.000001278984,0.08643854,0.0002627087,0.00002444511,0.000001293062,4.978316e-7,0.000004221466,0.005805823],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9177197,"threshold_uncertainty_score":0.3492547,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1460175943364272,"score_gpt":0.2942268895039399,"score_spread":0.1482092951675127,"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."}}