{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00006847917,0.0001613564,0.0001813188,0.0001647714,0.0003991607,0.00003423368,0.0002207558,0.00003270925,0.0002682197],"category_scores_gemma":[0.000001620883,0.0001569244,0.00008537165,0.0006247821,0.0001086205,0.0003041328,0.00007255768,0.0002358708,0.00003893855],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00008187988,"about_ca_system_score_gemma":0.0001309062,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001928504,"about_ca_topic_score_gemma":0.000005048212,"domain_scores_codex":[0.9991555,0.00004861495,0.0001542606,0.000213104,0.0001812521,0.0002472314],"domain_scores_gemma":[0.9995055,0.00006024019,0.00006991772,0.0001455205,0.0001280716,0.00009080377],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001828714,0.0001468497,0.003911739,0.00003926971,0.0002331984,0.000006331642,0.0002939511,0.2315049,0.0001100956,0.7564492,0.0003210042,0.006800607],"study_design_scores_gemma":[0.03111133,0.002554816,0.01931253,0.001939644,0.002691597,0.000008999677,0.02244957,0.4029431,0.01746269,0.2351842,0.2589799,0.005361645],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1194167,0.000002995077,0.7683051,0.0002073856,0.00004001831,0.0001679473,0.000004296213,0.00004965337,0.111806],"genre_scores_gemma":[0.9955018,0.000002676034,0.0005416048,0.00007495753,0.0000456919,0.000002553543,0.0000787143,0.000008331461,0.003743717],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.876085,"threshold_uncertainty_score":0.639919,"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."}}