{"id":"W2597452529","doi":"10.48550/arxiv.1703.04782","title":"Online Learning Rate Adaptation with Hypergradient Descent","year":2017,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Stochastic Gradient Optimization Techniques","field":"Computer Science","cited_by":77,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Computer science; Stochastic gradient descent; Gradient descent; Rate of convergence; Range (aeronautics); Computation; Convergence (economics); Adaptation (eye); Online machine learning; Descent (aeronautics); Mode (computer interface); Artificial intelligence; Mathematical optimization; Machine learning; Algorithm; Active learning (machine learning); Mathematics; Artificial neural network; Key (lock)","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":[0.0002250457,0.0003200913,0.0002888842,0.0002985101,0.0003481787,0.000228955,0.001690134,0.000181213,0.00000761917],"category_scores_gemma":[0.00008785023,0.0003332429,0.0001051042,0.000265114,0.0001269188,0.0004057326,0.001146629,0.0005909484,0.00001802111],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002549296,"about_ca_system_score_gemma":0.0001887902,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001329374,"about_ca_topic_score_gemma":0.0000525245,"domain_scores_codex":[0.998225,0.0001380644,0.0001753384,0.001026999,0.0001167675,0.0003178484],"domain_scores_gemma":[0.9976943,0.00006836835,0.0005202572,0.001243565,0.0003156232,0.0001579612],"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":[0.00002551047,0.0001186883,0.0008947054,0.00002868877,0.0000612951,0.0001468234,0.00033046,0.912534,0.000007518436,0.08421464,0.00004148088,0.001596187],"study_design_scores_gemma":[0.0003926331,0.0001602288,0.001568159,0.000205117,0.00006521183,0.000007233202,0.00006192622,0.9862784,0.00007932344,0.01056088,0.000209011,0.0004118532],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.06091542,0.00003321239,0.9369618,0.0001163403,0.0003029696,0.0003523991,0.00000627147,0.0006201742,0.0006914337],"genre_scores_gemma":[0.9497052,0.000136506,0.04877264,0.00003823432,0.00004082975,0.000002253217,0.00004922116,0.00002384554,0.001231235],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8887898,"threshold_uncertainty_score":0.999912,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08496107352918601,"score_gpt":0.1946122829403172,"score_spread":0.1096512094111312,"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."}}