{"id":"W6891697460","doi":"10.48448/c49p-n242","title":"Challenges in Applying Explainability Methods to Improve the Fairness of NLP Models","year":2022,"lang":"en","type":"other","venue":"Underline Science Inc.","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"National Research Council Canada","funders":"","keywords":"Training set; Interpretability; Key (lock); Business intelligence","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.01468489,0.0004919665,0.0007749022,0.00168808,0.0002710803,0.00006860511,0.003371767,0.000187483,0.0007812806],"category_scores_gemma":[0.00103933,0.0003716479,0.0001102494,0.00328879,0.001634191,0.000309644,0.00197764,0.0007451201,0.00006856955],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001043331,"about_ca_system_score_gemma":0.0009998203,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00175147,"about_ca_topic_score_gemma":0.003698753,"domain_scores_codex":[0.9941908,0.0008478737,0.0006932067,0.0016931,0.001658582,0.0009164219],"domain_scores_gemma":[0.9962193,0.0005141337,0.0005529404,0.002318392,0.0002012238,0.0001940614],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0001054607,0.001027927,0.0001712762,0.0007613857,0.00007955416,0.00001690722,0.0127738,0.02367113,0.01832242,0.1120709,0.006602278,0.8243969],"study_design_scores_gemma":[0.002160854,0.0009260295,0.0006327218,0.0006210324,0.0001745116,0.00003540768,0.07669042,0.2853945,0.00546264,0.14379,0.4803395,0.003772436],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"other","genre_gemma":"methods","genre_scores_codex":[0.0008148187,0.006231444,0.05838217,0.001982168,0.00159931,0.008746437,0.000451972,0.0007230623,0.9210686],"genre_scores_gemma":[0.3817154,0.0007167137,0.5403892,0.0006697254,0.0008825765,0.008940784,0.0000805941,0.003821365,0.06278361],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.858285,"threshold_uncertainty_score":0.9998735,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1271227836595832,"score_gpt":0.3973989403663739,"score_spread":0.2702761567067907,"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."}}