{"id":"W2581624817","doi":"10.48550/arxiv.1701.06538","title":"Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer","year":2017,"lang":"en","type":"article","venue":"arXiv (Cornell University)","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":268,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Institute for Catastrophic Loss Reduction","keywords":"Computer science; Computation; Machine translation; Artificial neural network; Language model; Artificial intelligence; Layer (electronics); Translation (biology); Machine learning; Language translation; Algorithm","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001381674,0.0001732252,0.0001738458,0.00004612949,0.0008411005,0.0001060214,0.002713661,0.00007824716,0.00001343223],"category_scores_gemma":[0.00003220516,0.0001434971,0.0001150094,0.0003583329,0.0001897682,0.000684171,0.0006740626,0.000204025,0.00002297332],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002947714,"about_ca_system_score_gemma":0.00002027254,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003351,"about_ca_topic_score_gemma":0.0000876814,"domain_scores_codex":[0.99879,0.00007448136,0.000147984,0.0005139583,0.000075596,0.000397943],"domain_scores_gemma":[0.9974241,0.0001405803,0.0002828275,0.001941822,0.0001012801,0.0001094258],"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":[0.00004131392,0.0001770348,0.01174335,0.000007194967,0.000064531,0.0002335536,0.0003067732,0.4431343,0.0003658707,0.5373931,0.003236788,0.003296176],"study_design_scores_gemma":[0.0004145,0.00003306386,0.004797014,0.00000911787,0.00001852636,0.00001009006,0.00002936941,0.9827024,0.0003070243,0.005385359,0.006086052,0.0002074448],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2065127,0.00006583061,0.7905393,0.0007792255,0.0003242675,0.000258607,0.000003869408,0.0001620659,0.001354145],"genre_scores_gemma":[0.9980142,0.00006456624,0.0006327114,0.0003093615,0.00008744022,0.000001048037,0.000002387172,0.00001241332,0.0008759386],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7915015,"threshold_uncertainty_score":0.646915,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0848082490162567,"score_gpt":0.2135652458888167,"score_spread":0.12875699687256,"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."}}