{"id":"W4382466676","doi":"10.1609/aaai.v37i7.25997","title":"Inverse-Reference Priors for Fisher Regularization of Bayesian Neural Networks","year":2023,"lang":"en","type":"article","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","topic":"Fault Detection and Control Systems","field":"Engineering","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kootenay Association for Science & Technology","funders":"Ministry of Science and ICT, South Korea; National Research Foundation of Korea; National Research Foundation","keywords":"Prior probability; Bayesian probability; Generalization; Fisher information; Regularization (linguistics); Inverse problem; Computer science; Inverse; Artificial neural network; Artificial intelligence; Benchmark (surveying); Deep neural networks; Machine learning; Mathematics","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00456376,0.001094793,0.0008397796,0.001298315,0.0005945275,0.001287901,0.001775476,0.001743862,0.002678677],"category_scores_gemma":[0.01662946,0.0006421118,0.0008415651,0.0009827516,0.001777299,0.002413822,0.001644172,0.003022294,0.0006697432],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001470433,"about_ca_system_score_gemma":0.001215997,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00410737,"about_ca_topic_score_gemma":0.003941494,"domain_scores_codex":[0.9987549,0.000488707,0.00005604146,0.0002418352,0.0003933637,0.00006515925],"domain_scores_gemma":[0.9962524,0.002424492,0.0002685098,0.0004497851,0.0005107494,0.00009402939],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0000681747,0.00004895644,0.0005793816,0.000124497,0.00004720063,0.00007444542,0.0001049405,0.6644313,0.003670717,0.2637604,0.00311825,0.0639717],"study_design_scores_gemma":[0.000004924988,0.000009239885,0.0001427928,0.00001873213,0.000004070796,0.00001868211,0.0000057721,0.9230256,0.0006671692,0.07499394,0.001097273,0.00001187697],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.003071861,0.000182437,0.9946667,0.0001986543,0.00001656227,0.00001917967,0.00006423629,0.0001516933,0.001628718],"genre_scores_gemma":[0.337782,0.001378779,0.653056,0.0006176411,0.0002467906,0.0004556522,0.0009421415,0.0004397815,0.005081239],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.00456376,"threshold_uncertainty_score":0.02413577,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0468949273352039,"score_gpt":0.2601142447525476,"score_spread":0.2132193174173437,"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."}}