{"id":"W4382317969","doi":"10.1609/aaai.v37i12.26779","title":"DeepGemini: Verifying Dependency Fairness for Deep Neural Network","year":2023,"lang":"en","type":"article","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"Japan Society for the Promotion of Science; Canada First Research Excellence Fund; JST-Mirai Program; University of Alberta; Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research","keywords":"Computer science; Counterexample; Heuristics; Certification; Scalability; Deep neural networks; Benchmark (surveying); Key (lock); Fairness measure; Artificial neural network; Artificial intelligence; Computer security; Mathematics","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.009021072,0.001723532,0.001248558,0.001301794,0.001358617,0.002192729,0.003892409,0.001799967,0.003929485],"category_scores_gemma":[0.04352696,0.000952695,0.002230858,0.0006893822,0.003537555,0.004657349,0.004852466,0.005242491,0.0005160307],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.005049775,"about_ca_system_score_gemma":0.008002029,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008609929,"about_ca_topic_score_gemma":0.01408064,"domain_scores_codex":[0.9925171,0.002356501,0.0004904558,0.001612878,0.002210075,0.0008131139],"domain_scores_gemma":[0.9774427,0.01472208,0.001529098,0.003405075,0.002285078,0.0006158897],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0009478742,0.0002250615,0.009682186,0.0005283565,0.0001851062,0.0002933627,0.0002794238,0.7413438,0.01012705,0.1034519,0.005102355,0.1278335],"study_design_scores_gemma":[0.00003170296,0.0000576167,0.0002068484,0.00002683743,0.00001814621,0.00003180519,0.00001803094,0.9477547,0.00523602,0.04583604,0.0007696762,0.0000125865],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0501229,0.0005751514,0.9390872,0.0008616158,0.0001721962,0.0001969287,0.0005684851,0.00570887,0.00270672],"genre_scores_gemma":[0.7480003,0.0003203117,0.2463535,0.0008316457,0.0001033396,0.0003872656,0.001054143,0.0006864857,0.002262974],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.009021072,"threshold_uncertainty_score":0.04770857,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08482360610997083,"score_gpt":0.3193312691216048,"score_spread":0.2345076630116339,"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."}}