{"id":"W2964309400","doi":"","title":"Revisiting Natural Gradient for Deep Networks","year":2014,"lang":"en","type":"article","venue":"","topic":"Neural Networks and Applications","field":"Computer Science","cited_by":139,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"","keywords":"Hessian matrix; Stochastic gradient descent; Gradient descent; Robustness (evolution); Diagonal; Generalization; Computer science; Metric (unit); Deep learning; Subspace topology; Algorithm; Artificial intelligence; Mathematics; Mathematical optimization; Applied mathematics; Artificial neural network","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":[0.0001407616,0.0000586492,0.00006753112,0.00001305457,0.000144577,0.00009534238,0.0003589569,0.00001764446,0.000002808705],"category_scores_gemma":[0.00001160729,0.00004520708,0.00004936519,0.0001357135,0.000009188034,0.0001088658,0.00007889166,0.00005680432,0.000007779307],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000006683963,"about_ca_system_score_gemma":0.000002242785,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000002157411,"about_ca_topic_score_gemma":0.00000314649,"domain_scores_codex":[0.9994389,0.00001243453,0.000104462,0.0001990109,0.00005270309,0.0001925078],"domain_scores_gemma":[0.9994796,0.0001433853,0.0000357003,0.0002587321,0.00003469567,0.00004788905],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[4.403445e-7,0.000003685957,0.00003439907,0.000002212743,0.000001733865,7.877299e-8,0.000008640781,0.002581548,0.00003985174,0.6635963,0.001871224,0.3318599],"study_design_scores_gemma":[0.00007105897,0.00001151424,0.0003066238,0.000004593229,0.000001435435,0.000002522215,0.000001312303,0.9607717,0.00006154161,0.004908659,0.03378922,0.0000698352],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001089396,0.0001057442,0.9942438,0.002170339,0.0002145439,0.0001373109,7.302562e-8,0.0001274151,0.001911362],"genre_scores_gemma":[0.9050089,0.000005772838,0.09275375,0.001387386,0.0004634897,0.00003644667,0.000002158488,0.000004048515,0.0003380329],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9581901,"threshold_uncertainty_score":0.1843491,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.009799498980287287,"score_gpt":0.2366033906371499,"score_spread":0.2268038916568626,"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."}}