{"id":"W4408566453","doi":"10.1016/j.knosys.2025.113318","title":"Mitigating covariance overfitting in out-of-distribution detection through intrinsic parameter learning","year":2025,"lang":"en","type":"article","venue":"Knowledge-Based Systems","topic":"Anomaly Detection Techniques and Applications","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of British Columbia","funders":"","keywords":"Overfitting; Covariance; Artificial intelligence; Computer science; Distribution (mathematics); Machine learning; Mathematics; Statistics; Artificial neural network; Mathematical analysis","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":[0.0006314307,0.0001446319,0.0002511826,0.0001495413,0.0002060798,0.0001050896,0.0003273402,0.000130721,0.000001409628],"category_scores_gemma":[0.0002484143,0.0001548442,0.00008402368,0.001256981,0.00004743545,0.0002434933,0.00008495832,0.000292643,0.0000169065],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002381139,"about_ca_system_score_gemma":0.0001257402,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001741758,"about_ca_topic_score_gemma":0.0000482426,"domain_scores_codex":[0.9984447,0.0002363363,0.0005680496,0.0003944434,0.0001250528,0.0002313701],"domain_scores_gemma":[0.9987245,0.0003749912,0.0002803642,0.0003871252,0.0002038638,0.00002916603],"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":[0.00008257793,0.0009293883,0.02328694,0.002376148,0.0001323545,0.00001200411,0.003714503,0.0318665,0.1055511,0.3487485,0.001046506,0.4822534],"study_design_scores_gemma":[0.0008077015,0.000148583,0.002497228,0.001070991,0.00001558964,0.000003446365,0.0002371173,0.7689433,0.2008777,0.002286857,0.02275887,0.0003526948],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02665298,0.000330328,0.9696409,0.00007000536,0.0006996081,0.0004009511,0.00000292068,0.0003255299,0.001876743],"genre_scores_gemma":[0.9920336,0.000004495609,0.007442723,0.00002092503,0.00005550348,0.0002160673,0.00000679913,0.000007843639,0.0002120214],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9653807,"threshold_uncertainty_score":0.6314363,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01820109285478196,"score_gpt":0.2818398798960486,"score_spread":0.2636387870412666,"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."}}