{"id":"W7124304825","doi":"10.65109/vzns8734","title":"A Model-Based Solution to the Offline Multi-Agent Reinforcement Learning Coordination Problem","year":2024,"lang":"","type":"article","venue":"","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Mila - Quebec Artificial Intelligence Institute","funders":"","keywords":"Reinforcement learning; Leverage (statistics); Observability; Online and offline; Simple (philosophy); Offline learning; Online algorithm","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":["metaepi_narrow","scholarly_communication","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.002547364,0.0005882421,0.0003458372,0.0005177119,0.001009231,0.002129808,0.00147289,0.0002042178,0.0002244708],"category_scores_gemma":[0.0002499573,0.0004504245,0.0002873554,0.001852005,0.00010458,0.0008767156,0.0009646273,0.001032537,0.001716304],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008979563,"about_ca_system_score_gemma":0.0008042773,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00009935995,"about_ca_topic_score_gemma":0.0000286635,"domain_scores_codex":[0.9947692,0.0002910448,0.001208108,0.001172192,0.00148004,0.001079377],"domain_scores_gemma":[0.9976558,0.0002202178,0.0002647381,0.001052483,0.0004886387,0.0003181105],"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":[0.00001908298,0.00005430624,0.00002175546,0.0002325381,0.00007691906,0.000008526648,0.003151089,0.946722,0.0006529949,0.01725327,0.005333897,0.02647356],"study_design_scores_gemma":[0.0005668162,0.0006796095,0.00001957204,0.0005294704,0.00008358627,0.000004885373,0.00008033254,0.9591917,0.0005298116,0.00003564804,0.03772355,0.0005549978],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.00007893245,0.0003207249,0.9676239,0.02383574,0.001596004,0.00216128,0.000001165403,0.0007307736,0.003651424],"genre_scores_gemma":[0.7995208,0.0000387921,0.1288564,0.001365054,0.0001883166,0.0001618233,0.00002948317,0.00005497696,0.06978438],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8387676,"threshold_uncertainty_score":0.9997947,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04538121631456204,"score_gpt":0.2906629308777778,"score_spread":0.2452817145632157,"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."}}