{"id":"W4412979458","doi":"10.1088/2632-2153/adf7fe","title":"Uncertainty quantification from ensemble variance scaling laws in deep neural networks","year":2025,"lang":"en","type":"article","venue":"Machine Learning Science and Technology","topic":"Anomaly Detection Techniques and Applications","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; University of Toronto","funders":"High Energy Physics","keywords":"Variance (accounting); Artificial neural network; Scaling law; Deep neural networks; Scaling; Law; Computer science; Artificial intelligence; Econometrics; Statistical physics; Mathematics; Political science; Economics; Physics; Accounting","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0005476082,0.00009472624,0.0001266231,0.000519169,0.0004901466,0.0001502212,0.0007749588,0.0001076457,0.00000244104],"category_scores_gemma":[0.0001801706,0.00009126846,0.00001385173,0.003610196,0.0003353279,0.0002514648,0.0003247258,0.0004178558,0.000002812659],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005500727,"about_ca_system_score_gemma":0.00004427089,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005061394,"about_ca_topic_score_gemma":0.0001554628,"domain_scores_codex":[0.9988462,0.00003278597,0.0001947129,0.0005425024,0.0001220764,0.0002616671],"domain_scores_gemma":[0.9993255,0.00007124813,0.00008156611,0.0003889246,0.0001028123,0.00002992997],"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":[0.000002832067,0.00002783305,0.02070154,0.000002812301,0.000002377778,0.000002372407,0.00007130235,0.01871726,0.009782397,0.3173532,0.000008990554,0.633327],"study_design_scores_gemma":[0.00009680323,0.00002371468,0.004746343,0.00001043548,0.00000182295,0.000003937427,0.00003513305,0.9715163,0.002029135,0.01976998,0.001678404,0.0000880358],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1032464,0.0003372291,0.8903908,0.004904571,0.00007959834,0.0001193059,2.878288e-7,0.0004641859,0.0004576073],"genre_scores_gemma":[0.9846098,0.00004102948,0.01501877,0.0002040718,0.000008953641,0.00004287402,0.000001706371,0.000003056261,0.00006976498],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.952799,"threshold_uncertainty_score":0.3769861,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.006595860381345152,"score_gpt":0.2530018814771641,"score_spread":0.246406021095819,"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."}}