{"id":"W2965050397","doi":"10.24963/ijcai.2019/434","title":"On Principled Entropy Exploration in Policy Optimization","year":2019,"lang":"en","type":"article","venue":"","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":18,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Benchmark (surveying); Entropy (arrow of time); Mathematical optimization; Optimization problem; Convergence (economics); Set (abstract data type); Monotonic function; Constrained optimization problem; Multi-objective optimization; Descent (aeronautics); Machine learning; Algorithm; Mathematics; Economics; Engineering","routes":{"ca_aff":true,"ca_fund":true,"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":[],"consensus_categories":[],"category_scores_codex":[0.0001189901,0.00006423348,0.00006545916,0.0002019105,0.00001885267,0.0000905363,0.0002777955,0.00003158208,0.00008496971],"category_scores_gemma":[0.00007346177,0.0000578885,0.00001554239,0.00037813,0.000004411512,0.0006263934,0.00008157518,0.00006768735,0.0004793944],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00009722939,"about_ca_system_score_gemma":0.00005027052,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001644904,"about_ca_topic_score_gemma":0.000001240219,"domain_scores_codex":[0.9993034,0.00003475381,0.0001490681,0.0001724389,0.0001974424,0.0001429259],"domain_scores_gemma":[0.9995272,0.00004362499,0.00005354339,0.0003219044,0.0000275562,0.00002619403],"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":[0.000002051983,0.000007713348,0.0004454385,0.000001643847,7.853649e-7,3.400396e-7,0.000128948,0.707031,0.00004874916,0.2921036,0.00002248409,0.000207189],"study_design_scores_gemma":[0.0003188381,0.00009341364,0.0004661588,0.00000860919,2.469181e-7,3.000674e-7,0.000008255121,0.9980679,0.0003226116,0.0003897412,0.000255807,0.00006807403],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.003435982,6.649079e-7,0.9735467,0.0009968863,0.0001394517,0.0001960214,2.619299e-8,0.0001002516,0.02158404],"genre_scores_gemma":[0.904057,0.000007045667,0.09022417,0.000617407,0.00002981906,0.000008014737,0.000005080038,0.000006448655,0.005045039],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.900621,"threshold_uncertainty_score":0.6161805,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01397620461221087,"score_gpt":0.254103345513721,"score_spread":0.2401271409015101,"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."}}