{"id":"W4388514504","doi":"10.48550/arxiv.2311.03683","title":"Preventing Arbitrarily High Confidence on Far-Away Data in Point-Estimated Discriminative Neural Networks","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Government of Canada; Canadian Institute for Advanced Research","keywords":"Discriminative model; Computer science; Artificial neural network; Training set; Class (philosophy); Machine learning; Test data; Artificial intelligence; Logit; De facto; Point (geometry); Set (abstract data type); Data set; Mathematics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01178196,0.001892816,0.002154844,0.0009409169,0.0009912311,0.001416786,0.003709556,0.003209392,0.001791909],"category_scores_gemma":[0.0617891,0.001257342,0.00117611,0.0008932988,0.004524347,0.004170937,0.009776874,0.006585712,0.0009910219],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001576671,"about_ca_system_score_gemma":0.00093607,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002121399,"about_ca_topic_score_gemma":0.002521483,"domain_scores_codex":[0.9919232,0.003678192,0.0004085301,0.001438477,0.002079157,0.0004725027],"domain_scores_gemma":[0.9485549,0.03902144,0.002519372,0.00692885,0.002243705,0.0007318258],"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":[0.001146132,0.0002275291,0.006505988,0.0003554973,0.0001570623,0.0005688432,0.0003323437,0.8227345,0.00996754,0.02456626,0.005413555,0.1280247],"study_design_scores_gemma":[0.00004021025,0.0001327566,0.0007218926,0.00006819217,0.0000148693,0.0002181689,0.00003762304,0.9709882,0.00743331,0.01958418,0.0007314525,0.0000291845],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09174901,0.00111245,0.9006879,0.001229382,0.0001142609,0.00008567253,0.0002536402,0.001800401,0.002967305],"genre_scores_gemma":[0.8770046,0.0003967191,0.1164874,0.001174384,0.0001366459,0.0001980736,0.0008543857,0.0004151438,0.003332672],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01178196,"threshold_uncertainty_score":0.06230968,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1618378683921594,"score_gpt":0.2594925872341089,"score_spread":0.09765471884194954,"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."}}