{"id":"W2963344792","doi":"","title":"Fast Approximate Natural Gradient Descent in a Kronecker Factored Eigenbasis","year":2018,"lang":"en","type":"article","venue":"Neural Information Processing Systems","topic":"Face and Expression Recognition","field":"Computer Science","cited_by":45,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"","keywords":"Diagonal; Kronecker delta; Leverage (statistics); Gradient descent; Computer science; Covariance; Stochastic gradient descent; Algorithm; Mathematics; Mathematical optimization; Artificial intelligence; Artificial neural network; Statistics; Physics; Geometry","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":[0.001235811,0.0009464217,0.001092157,0.0004820971,0.0003843441,0.001172647,0.00097537,0.001268015,0.003689656],"category_scores_gemma":[0.005300466,0.0005737891,0.0005406461,0.0006023757,0.0008748181,0.001747174,0.0009394187,0.001564235,0.001774415],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007713774,"about_ca_system_score_gemma":0.001638914,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006117841,"about_ca_topic_score_gemma":0.01008113,"domain_scores_codex":[0.9995176,0.0001446242,0.00002917457,0.0001078391,0.000150727,0.00005011033],"domain_scores_gemma":[0.998995,0.0003381471,0.00006615913,0.0002356967,0.0003110487,0.00005383101],"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":[0.0001689771,0.0001089828,0.001104967,0.0001251505,0.0000721544,0.00008883585,0.0001369081,0.7415743,0.007732449,0.06457617,0.009242476,0.1750687],"study_design_scores_gemma":[0.000007561085,0.00001348488,0.00006706859,0.000005544558,0.000002439076,0.00001403598,0.000005658133,0.9910623,0.0004812407,0.007610166,0.0007260356,0.00000447284],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.006306186,0.00008472524,0.991864,0.00008073637,0.00004021765,0.00002349307,0.00003144857,0.00062038,0.0009488334],"genre_scores_gemma":[0.2293572,0.0001739565,0.7646148,0.0001975409,0.00005083096,0.0001664843,0.0002814615,0.0004138627,0.004743783],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006117841,"threshold_uncertainty_score":0.01234311,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01720553793103738,"score_gpt":0.2423449592092438,"score_spread":0.2251394212782065,"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."}}