{"id":"W2155894447","doi":"10.48550/arxiv.1503.05671","title":"Optimizing Neural Networks with Kronecker-factored Approximate Curvature","year":2015,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Neural Networks and Applications","field":"Computer Science","cited_by":292,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Kronecker delta; Curvature; Stochastic gradient descent; Hessian matrix; Diagonal; Gradient descent; Artificial neural network; Mathematics; Matrix (chemical analysis); Invertible matrix; Applied mathematics; Rank (graph theory); Mathematical optimization; Algorithm; Computer science; Combinatorics; Geometry; Pure mathematics; Artificial intelligence; Physics","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0001537566,0.0004371802,0.0003904653,0.0001167509,0.0002142015,0.0002931636,0.002109493,0.0003325283,0.000008768725],"category_scores_gemma":[0.000004899181,0.0004069874,0.0001465271,0.000789118,0.0001221629,0.0004645302,0.001853135,0.001103782,0.00001449261],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00013553,"about_ca_system_score_gemma":0.0001201356,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004565714,"about_ca_topic_score_gemma":0.0000305064,"domain_scores_codex":[0.9977612,0.00009407709,0.0001903725,0.001295348,0.0001243545,0.0005346385],"domain_scores_gemma":[0.9975487,0.00005691121,0.0003111738,0.001547085,0.0002147628,0.0003213817],"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":[0.00002358265,0.00004587022,0.0003644285,0.00001480974,0.00005134835,0.0001303051,0.0000890822,0.9774745,0.000003619603,0.02019645,0.0009479504,0.0006581317],"study_design_scores_gemma":[0.0003668276,0.00004900935,0.0001307788,0.00005564647,0.00005945777,0.00001115997,0.00002269734,0.9938928,0.00001129246,0.004106327,0.0007873484,0.0005066849],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.07142419,0.0002014507,0.9256894,0.000261731,0.0004104286,0.0004641962,0.00001189602,0.0003741727,0.001162509],"genre_scores_gemma":[0.9869789,0.00007948003,0.01199544,0.0001400977,0.00019401,0.000003801666,0.00005219498,0.00003158503,0.0005244248],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9155548,"threshold_uncertainty_score":0.9998382,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06471831599879245,"score_gpt":0.1881355193432239,"score_spread":0.1234172033444314,"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."}}