{"id":"W4381053704","doi":"10.1007/s11432-022-3580-5","title":"Toward actionable testing of deep learning models","year":2023,"lang":"en","type":"article","venue":"Science China Information Sciences","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Deep learning; Artificial intelligence; Data science","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":[0.01060179,0.001799322,0.001659909,0.001380153,0.0006498849,0.002040582,0.004226581,0.003653403,0.003800986],"category_scores_gemma":[0.05678527,0.001067825,0.00146375,0.0006188178,0.00423633,0.004283684,0.006756356,0.004851622,0.0004975392],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002167496,"about_ca_system_score_gemma":0.002465585,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002721635,"about_ca_topic_score_gemma":0.002660771,"domain_scores_codex":[0.9926162,0.003879616,0.0003495215,0.001106318,0.001454168,0.0005942182],"domain_scores_gemma":[0.9460824,0.04246091,0.002037492,0.00596123,0.002271336,0.001186601],"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":[0.0006230421,0.0001964498,0.003362044,0.0001727135,0.0001592635,0.0002695911,0.0001466127,0.8291138,0.005351017,0.0918229,0.003523693,0.06525893],"study_design_scores_gemma":[0.0000143834,0.00002576126,0.00008094108,0.00001109375,0.00000515261,0.00001121214,0.000006634336,0.9634984,0.001014864,0.03519889,0.0001281041,0.000004480868],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.04806603,0.0002280449,0.9447088,0.001602961,0.00008601153,0.00007734127,0.0002219148,0.002130653,0.002878146],"genre_scores_gemma":[0.8677555,0.00009717879,0.1278767,0.0006273332,0.00009226659,0.0001713889,0.0005565269,0.0003715123,0.002451579],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01060179,"threshold_uncertainty_score":0.0560683,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06458478689624302,"score_gpt":0.3011856642803998,"score_spread":0.2366008773841567,"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."}}