{"id":"W4386585729","doi":"10.1007/978-3-031-40923-3_7","title":"A Low-Cost Strategic Monitoring Approach for Scalable and Interpretable Error Detection in Deep Neural Networks","year":2023,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":8,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of British Columbia","funders":"Natural Sciences and Engineering Research Council of Canada; European Commission","keywords":"Computer science; Anomaly detection; Inference; Scalability; Overhead (engineering); Artificial intelligence; Categorization; Artificial neural network; Precision and recall; Recall; Component (thermodynamics); Fault detection and isolation; Detector; Pattern recognition (psychology); Machine learning; Data mining; Programming language","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":[0.001592742,0.001293312,0.001280078,0.0006018681,0.0004625681,0.001273191,0.002947479,0.00155911,0.003670445],"category_scores_gemma":[0.005623891,0.0006877548,0.0006612738,0.0006590698,0.0009005669,0.002223496,0.003713191,0.002808298,0.001072456],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008530689,"about_ca_system_score_gemma":0.001557132,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00268331,"about_ca_topic_score_gemma":0.004895897,"domain_scores_codex":[0.9988808,0.0002325368,0.00006817064,0.0002732899,0.0004159465,0.0001292906],"domain_scores_gemma":[0.9980141,0.0008617813,0.0001727732,0.0004434697,0.0003975127,0.0001103646],"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":[0.0003877549,0.000207185,0.0008869461,0.0001303225,0.0001017785,0.0001571033,0.0001081869,0.2763551,0.02142227,0.04456162,0.01045738,0.6452244],"study_design_scores_gemma":[0.000006008555,0.00002772841,0.00007097457,0.000005651528,0.000007116118,0.00002128045,0.000005147164,0.9840819,0.002828361,0.01240275,0.0005362399,0.00000673235],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.004438404,0.000185818,0.9931307,0.0001462589,0.00006713453,0.0000341729,0.00005615272,0.001048105,0.0008931572],"genre_scores_gemma":[0.3817094,0.0002952186,0.6085647,0.0003746482,0.000200283,0.0001907432,0.0004408801,0.0003533908,0.007870812],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003670445,"threshold_uncertainty_score":0.01227891,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02756579282268076,"score_gpt":0.266124434473667,"score_spread":0.2385586416509862,"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."}}