{"id":"W2996001434","doi":"10.48550/arxiv.1912.05128","title":"Marginalized State Distribution Entropy Regularization in Policy Optimization","year":2019,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Regularization (linguistics); Entropy (arrow of time); Mathematical optimization; Econometrics; Mathematics; Economics; Computer science; Statistical physics; Mathematical economics; Political science; Physics; Artificial intelligence; Thermodynamics","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":[0.00249518,0.001031873,0.001376798,0.0006674499,0.000442923,0.001153255,0.001013924,0.00132987,0.001947738],"category_scores_gemma":[0.01091154,0.0006021076,0.0007156251,0.0007391103,0.001946818,0.002002781,0.001742819,0.002289037,0.000368722],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001486814,"about_ca_system_score_gemma":0.001377395,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00236535,"about_ca_topic_score_gemma":0.001710516,"domain_scores_codex":[0.9984354,0.0008087091,0.0000696953,0.0002396131,0.0003229505,0.0001236089],"domain_scores_gemma":[0.9962321,0.002802345,0.0002386295,0.0002949861,0.0003072739,0.0001247359],"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":[0.00009931283,0.00006633869,0.0006866781,0.00008396856,0.0000571529,0.00004972496,0.00006876883,0.9212264,0.001935714,0.04392098,0.0009943589,0.03081065],"study_design_scores_gemma":[0.000007336259,0.00001838058,0.00007614457,0.000007348843,0.000004079841,0.000007938231,0.000002898713,0.9804863,0.0004584611,0.01868899,0.0002372384,0.000004946665],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01113789,0.000318556,0.9865829,0.0002789458,0.00003896717,0.00002727743,0.00003063954,0.000204202,0.001380472],"genre_scores_gemma":[0.8629543,0.0005522032,0.1320477,0.0003344502,0.0001535742,0.0002222925,0.0001625562,0.0002092573,0.00336365],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.00249518,"threshold_uncertainty_score":0.01319593,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03583607267643475,"score_gpt":0.1904179922911267,"score_spread":0.1545819196146919,"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."}}