{"id":"W2963939004","doi":"","title":"Flipout: Efficient Pseudo-Independent Weight Perturbations on Mini-Batches","year":2018,"lang":"en","type":"article","venue":"arXiv (Cornell University)","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":20,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Variance reduction; Computer science; Multiplicative function; Control variates; Gaussian; Stochastic gradient descent; Regularization (linguistics); Importance sampling; Convolutional neural network; Artificial neural network; Reduction (mathematics); Algorithm; Variance (accounting); Bayesian probability; Artificial intelligence; Machine learning; Monte Carlo method; Mathematics; Markov chain Monte Carlo; Statistics","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.00008971986,0.0001841152,0.0001328576,0.0001541658,0.00044288,0.00005127405,0.001094329,0.00007266513,0.00005464814],"category_scores_gemma":[0.0000233659,0.0001894448,0.00008909326,0.001052539,0.0001703116,0.0002499263,0.0002804524,0.0001638067,0.00102554],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000140609,"about_ca_system_score_gemma":0.00005155979,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000009143363,"about_ca_topic_score_gemma":0.00002144315,"domain_scores_codex":[0.9985633,0.00005629825,0.0001399625,0.0007708526,0.0001191254,0.0003504883],"domain_scores_gemma":[0.9984027,0.0001874076,0.0001049185,0.0009871589,0.0001406778,0.0001771946],"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":[0.00002498278,0.0002535146,0.0006536093,0.000003619865,0.00002546952,0.00005124721,0.0004127034,0.06896191,0.0007762123,0.9248751,0.001803638,0.002157937],"study_design_scores_gemma":[0.0006283547,0.0002456312,0.0042848,0.00002325381,0.00002657999,0.00001267707,0.00008018706,0.9588282,0.003662233,0.02061254,0.01113083,0.0004647334],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3431466,0.000009194706,0.6481718,0.0006168276,0.000300497,0.0002146594,0.000004008946,0.0002758839,0.007260508],"genre_scores_gemma":[0.9919227,0.00001587731,0.004036958,0.0003954953,0.0001664386,0.000002330111,0.000002941887,0.00001255525,0.003444652],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9042626,"threshold_uncertainty_score":0.9997523,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04482411746879406,"score_gpt":0.1925983170088435,"score_spread":0.1477741995400494,"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."}}