{"id":"W7076000910","doi":"","title":"Reinforcement learning with non-ergodic reward increments: robustness via ergodicity transformations","year":2025,"lang":"en","type":"article","venue":"Aaltodoc (Aalto University)","topic":"Theoretical and Computational Physics","field":"Physics and Astronomy","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Cybernet Systems Corporation (Canada)","funders":"Engineering and Physical Sciences Research Council; Vetenskapsrådet; University of Warwick","keywords":"Reinforcement learning; Robustness (evolution); Ergodicity; Transformation (genetics); Probability distribution; Expected return; Expected value","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":[0.003645594,0.001077385,0.001251015,0.000747077,0.0004770997,0.001392513,0.001118634,0.001120909,0.001789944],"category_scores_gemma":[0.02059833,0.0005038156,0.0007675854,0.0003943274,0.002337486,0.00201283,0.002216225,0.002323816,0.0003167952],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001405038,"about_ca_system_score_gemma":0.0012059,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002570548,"about_ca_topic_score_gemma":0.001386424,"domain_scores_codex":[0.9987394,0.000502445,0.00007384326,0.0002726021,0.0002423757,0.0001693871],"domain_scores_gemma":[0.9889362,0.008485465,0.001078992,0.0006561622,0.0004625763,0.0003804525],"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":[0.000101752,0.00004831728,0.001128831,0.00003898595,0.0000486231,0.00007443128,0.00006633725,0.9483463,0.001311951,0.03601367,0.000429666,0.01239119],"study_design_scores_gemma":[0.00001144648,0.00003254019,0.0001035969,0.000005875861,0.000005152879,0.00001181727,0.000003938292,0.979615,0.0003491933,0.01972285,0.0001321826,0.000006410574],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.06257006,0.0003228928,0.931865,0.0005548186,0.00003898017,0.0000667487,0.00006845554,0.0005972578,0.003915826],"genre_scores_gemma":[0.9571691,0.0001796997,0.04045857,0.0001302982,0.00004182176,0.0001040848,0.00006671304,0.0001146863,0.001734915],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003645594,"threshold_uncertainty_score":0.01928002,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.002918719366986432,"score_gpt":0.1823896631469083,"score_spread":0.1794709437799219,"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."}}