{"id":"W3105702366","doi":"","title":"Variational Policy Gradient Method for Reinforcement Learning with General Utilities","year":2020,"lang":"en","type":"article","venue":"Neural Information Processing Systems","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Mathematical optimization; Reinforcement learning; Convexity; Markov decision process; Q-learning; Mathematics; Gradient descent; Gradient method; Computer science; Applied mathematics; Convergence (economics); Markov process; Artificial neural network; Artificial intelligence; Finance","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":[0.002291683,0.001116693,0.001346153,0.0007061209,0.0004145583,0.001031678,0.001399044,0.001493598,0.004050528],"category_scores_gemma":[0.00581927,0.0006667115,0.0008064877,0.00057924,0.001692649,0.001345704,0.001683992,0.002435208,0.0006902316],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002098375,"about_ca_system_score_gemma":0.002348468,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007117778,"about_ca_topic_score_gemma":0.004613396,"domain_scores_codex":[0.9993724,0.0003241123,0.0000219371,0.00008687683,0.0001398188,0.00005487389],"domain_scores_gemma":[0.9984523,0.001162686,0.00006995718,0.00006584479,0.0001840967,0.00006503834],"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":[0.00004076752,0.00003206255,0.0002944705,0.00009556292,0.00003928955,0.00006238296,0.00006446524,0.7716489,0.0008335746,0.2065126,0.001337097,0.01903883],"study_design_scores_gemma":[0.000005196491,0.000005668434,0.00001489698,0.000004670444,0.00000183202,0.000003657381,0.000002581621,0.9804669,0.00008627094,0.01888222,0.0005237366,0.000002384588],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.001873117,0.0002401397,0.9957187,0.0002153773,0.00003649751,0.00003389629,0.00001997638,0.0000860828,0.001776316],"genre_scores_gemma":[0.4195734,0.001070732,0.5589846,0.0004870342,0.0001817075,0.0007639742,0.0002646429,0.0004831968,0.01819076],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007117778,"threshold_uncertainty_score":0.01522487,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03067472875020712,"score_gpt":0.2824091564659542,"score_spread":0.2517344277157471,"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."}}