{"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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.001271707,0.0005335446,0.000571937,0.000885391,0.0003415768,0.0007952929,0.001909428,0.0004086646,0.000001493967],"category_scores_gemma":[0.0001386282,0.0005312923,0.00009896375,0.0009830474,0.0003415803,0.0008649231,0.001353502,0.001416612,0.000002363248],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003412688,"about_ca_system_score_gemma":0.0001241376,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000057589,"about_ca_topic_score_gemma":0.0001290389,"domain_scores_codex":[0.9962592,0.00006220421,0.0005288804,0.001715017,0.000527703,0.0009070213],"domain_scores_gemma":[0.9980268,0.0006691668,0.0002680225,0.0007555661,0.000131257,0.00014912],"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":[0.00001194431,0.000009529998,0.0001242331,0.0000513055,0.000004419056,0.00001607429,0.0002707852,0.7213129,0.00002638097,0.0008444939,1.541439e-7,0.2773277],"study_design_scores_gemma":[0.0003672176,0.00009828759,0.0001376265,0.0002446247,0.000007100267,0.00003190356,0.000001980464,0.9774721,0.0001392529,0.0209996,0.000005200673,0.000495091],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0004141293,0.0001642029,0.9954127,0.00006597788,0.002541454,0.0008477375,0.000001270377,0.0002289652,0.0003235415],"genre_scores_gemma":[0.757369,0.00001486031,0.2417695,0.0000820544,0.0005537164,0.00006770575,0.000004343098,0.00005736831,0.00008146595],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7569548,"threshold_uncertainty_score":0.9997138,"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."}}