{"id":"W4287865816","doi":"10.1115/1.4068705","title":"The Search for Feedback in Reinforcement Learning","year":2025,"lang":"en","type":"preprint","venue":"Journal of Dynamic Systems Measurement and Control","topic":"Neural Networks and Reservoir Computing","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Rockwell Automation (Canada)","funders":"","keywords":"Reinforcement learning; Parametrization (atmospheric modeling); Representation (politics); Computer science; Nonlinear system; Sequence (biology); State (computer science); Feedback loop; Open-loop controller; Control theory (sociology); Dynamical systems theory; Mathematical optimization; Closed loop; Artificial intelligence; Mathematics; Algorithm; Law; Control engineering; Engineering; Control (management)","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":[0.001379558,0.0005173768,0.001013728,0.000309348,0.0003932707,0.0009366321,0.0008572532,0.001262724,0.002172856],"category_scores_gemma":[0.007655892,0.0004060908,0.0002839643,0.0002603631,0.00178813,0.001275682,0.001382493,0.001471428,0.0002477763],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006590588,"about_ca_system_score_gemma":0.0006337986,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001829703,"about_ca_topic_score_gemma":0.0009740053,"domain_scores_codex":[0.9995134,0.0002344639,0.00002242269,0.00009023547,0.00007862301,0.00006084172],"domain_scores_gemma":[0.9967428,0.002539007,0.0002466755,0.0001460787,0.0002113647,0.0001140813],"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":[0.0001759713,0.0000528208,0.0007466326,0.0001185345,0.00005131391,0.0001067492,0.00008810483,0.9160005,0.003879901,0.05400414,0.0009380648,0.02383724],"study_design_scores_gemma":[0.0000118393,0.00002256319,0.00003988077,0.000006569207,0.00000314856,0.000006982116,0.000004385486,0.9842106,0.0003928922,0.01515437,0.0001435292,0.000003296149],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09655908,0.0008084803,0.8967623,0.001239792,0.00008111095,0.00003685347,0.00003985206,0.0004366168,0.004035834],"genre_scores_gemma":[0.9697868,0.0001556714,0.02825781,0.0001268354,0.00003757767,0.00005861289,0.00002153642,0.00004293862,0.001512237],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002172856,"threshold_uncertainty_score":0.007295907,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02384696423610495,"score_gpt":0.258297796052063,"score_spread":0.234450831815958,"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."}}