{"id":"W3160165418","doi":"10.1109/icse43902.2021.00044","title":"Self-Checking Deep Neural Networks in Deployment","year":2021,"lang":"en","type":"article","venue":"","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":36,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"National Satellite of Excellence in Trustworthy Software Systems, National University of Singapore; National Research Foundation Singapore; National University of Singapore; National Research Foundation","keywords":"Software deployment; Computer science; Artificial neural network; Artificial intelligence; Software engineering","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":[0.01004388,0.001441505,0.0006217065,0.0008211213,0.0005786317,0.002054193,0.002997373,0.001677261,0.002982956],"category_scores_gemma":[0.04938556,0.001364279,0.0006858069,0.0003873653,0.002233392,0.005935018,0.003125821,0.002942618,0.001566029],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001474614,"about_ca_system_score_gemma":0.001467658,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003742093,"about_ca_topic_score_gemma":0.003634199,"domain_scores_codex":[0.9928229,0.002533254,0.0004955623,0.001623868,0.0020255,0.0004989382],"domain_scores_gemma":[0.9674531,0.01494135,0.002273094,0.01148302,0.003235249,0.0006141441],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002724707,0.0005554747,0.04720613,0.0008241185,0.0004804932,0.002176822,0.001767251,0.2733772,0.07110507,0.03735009,0.07411307,0.4883195],"study_design_scores_gemma":[0.0001670503,0.0004784564,0.004502699,0.0002174438,0.00009100117,0.0007274122,0.0002403736,0.8780316,0.06974445,0.02633042,0.01934179,0.0001274054],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2920903,0.001915841,0.5637085,0.003321935,0.0009210497,0.0003413493,0.0005492684,0.127405,0.009746643],"genre_scores_gemma":[0.8752899,0.0002815925,0.1159223,0.001337142,0.00006566041,0.0001143303,0.0006988877,0.003381572,0.002908648],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01004388,"threshold_uncertainty_score":0.05311769,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0108506426513521,"score_gpt":0.2484945350835424,"score_spread":0.2376438924321903,"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."}}