{"id":"W4416429290","doi":"10.1109/tnnls.2025.3626050","title":"Universal Stabilization for Maximum Entropy Optimization in Reinforcement Learning","year":2025,"lang":"en","type":"article","venue":"IEEE Transactions on Neural Networks and Learning Systems","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"China Scholarship Council; National Natural Science Foundation of China","keywords":"Entropy (arrow of time); Kullback–Leibler divergence; Reinforcement learning; Principle of maximum entropy; Upper and lower bounds; Monotonic function; Maximum entropy spectral estimation; Divergence (linguistics)","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":[],"consensus_categories":[],"category_scores_codex":[0.0005233875,0.0001971421,0.000231612,0.0003268541,0.0005175739,0.0002986431,0.0002189867,0.0001281212,0.000004861368],"category_scores_gemma":[0.00002826368,0.0002040331,0.00005635128,0.0005965194,0.00003354825,0.0003985219,0.000006482359,0.0005967807,0.0000010405],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001433792,"about_ca_system_score_gemma":0.00003279385,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005878647,"about_ca_topic_score_gemma":0.000004855645,"domain_scores_codex":[0.9983131,0.0003431777,0.0004130312,0.0003931387,0.0001893362,0.0003482727],"domain_scores_gemma":[0.9990667,0.0004174136,0.0001649824,0.0001961153,0.00009485046,0.00006000788],"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":[0.00004073703,0.0000152676,0.0002366039,0.00005495566,0.00002169095,0.000001582117,0.0002148502,0.9935799,0.00002477828,0.001419063,0.0000218841,0.004368647],"study_design_scores_gemma":[0.0008387662,0.0003432804,0.00002965771,0.0001425531,0.00001699109,0.000002860759,0.0002198299,0.9971748,0.00002394887,0.000005103098,0.00102664,0.0001755475],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0009118175,0.00008890984,0.9966233,0.0001603448,0.001184314,0.0006374188,2.11476e-7,0.0002061326,0.0001875898],"genre_scores_gemma":[0.9952481,0.0001274302,0.001845468,0.00005863455,0.00003874432,0.00006258594,0.000009002853,0.00001673339,0.002593338],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9947778,"threshold_uncertainty_score":0.832023,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01099109538066441,"score_gpt":0.2301446614918618,"score_spread":0.2191535661111974,"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."}}