{"id":"W4402706509","doi":"10.48550/arxiv.2408.16262","title":"On Convergence of Average-Reward Q-Learning in Weakly Communicating Markov Decision Processes","year":2024,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Bayesian Modeling and Causal Inference","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; University of Alberta; DeepMind","keywords":"Markov decision process; Convergence (economics); Markov chain; Computer science; Artificial intelligence; Markov process; Mathematics; Psychology; Machine learning; Statistics; Economics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01306588,0.002227664,0.002752233,0.002210583,0.001284536,0.002709925,0.002800446,0.002652164,0.003519498],"category_scores_gemma":[0.06578741,0.001015631,0.002006956,0.001391347,0.005778942,0.005125304,0.004542337,0.004730232,0.0004883952],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003571315,"about_ca_system_score_gemma":0.00249218,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004783473,"about_ca_topic_score_gemma":0.001862484,"domain_scores_codex":[0.994755,0.002949334,0.0002326344,0.0007338304,0.000925402,0.0004036691],"domain_scores_gemma":[0.9171953,0.07438783,0.003037727,0.001312394,0.002852766,0.001214029],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0001524779,0.00008508106,0.001379939,0.0002185983,0.0001121942,0.0001038989,0.0003204063,0.6130358,0.0006990943,0.3708583,0.0008322002,0.01220197],"study_design_scores_gemma":[0.00001694411,0.00004314279,0.00011842,0.00003908142,0.00001168221,0.00001454095,0.00002125442,0.8844363,0.000244543,0.1147697,0.0002720379,0.00001241053],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03179121,0.001053374,0.9604083,0.001008954,0.00004096854,0.00008941007,0.00007935758,0.0001535287,0.005374924],"genre_scores_gemma":[0.8092283,0.002747481,0.1797384,0.0006552795,0.0002059354,0.0006641017,0.0003637773,0.0003285141,0.00606815],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01306588,"threshold_uncertainty_score":0.06909984,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06133400688008574,"score_gpt":0.2172719103401338,"score_spread":0.1559379034600481,"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."}}