{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001017901,0.000295149,0.0003435703,0.0001589189,0.000551446,0.0003215557,0.003194596,0.0001378787,0.00002537028],"category_scores_gemma":[0.001234604,0.0002418175,0.0001995961,0.001571713,0.0001942552,0.0006087926,0.000917937,0.0004552607,0.00008895597],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005194849,"about_ca_system_score_gemma":0.00007620457,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001851744,"about_ca_topic_score_gemma":0.00001127741,"domain_scores_codex":[0.9973667,0.00002986383,0.0006160085,0.0006889198,0.000581253,0.0007172236],"domain_scores_gemma":[0.9980455,0.000419516,0.0004645468,0.0003947655,0.0005697288,0.0001059616],"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":[0.00006862135,0.00003535183,0.0008357461,0.00006779698,0.00001831989,0.000001261235,0.001372025,0.02963696,0.001856478,0.8254738,0.00016378,0.1404698],"study_design_scores_gemma":[0.00003886927,0.0001130016,0.0003027231,0.000103438,0.00001340474,0.000003798736,0.0004497426,0.7479902,0.01542971,0.2352155,0.00009297601,0.0002466252],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09821355,0.00003773088,0.8867733,0.004576178,0.00313779,0.001169894,0.000004251022,0.0007222104,0.005365049],"genre_scores_gemma":[0.9856853,0.00001494458,0.01353938,0.0001649364,0.0003016158,0.00008863355,0.000001294224,0.00002850479,0.0001753482],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8874718,"threshold_uncertainty_score":0.986103,"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."}}