{"id":"W4416829986","doi":"10.48550/arxiv.2504.11258","title":"Multi-Agent Reinforcement Learning for Greenhouse Gas Offset Credit Markets","year":2025,"lang":"en","type":"preprint","venue":"ArXiv.org","topic":"Sustainable Finance and Green Bonds","field":"Economics, Econometrics and Finance","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Offset (computer science); Greenhouse gas; Nash equilibrium; Reinforcement learning; Carbon offset; Global warming; Control (management)","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.000945701,0.0005311084,0.0009630485,0.0005378176,0.0003198689,0.0001129385,0.0007162195,0.0005560807,0.0004042541],"category_scores_gemma":[0.0005523245,0.0006658806,0.0005336387,0.00026366,0.00005842798,0.0001775195,0.0009840695,0.000855039,0.0004648438],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006394803,"about_ca_system_score_gemma":0.0001959692,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007899699,"about_ca_topic_score_gemma":0.00007184909,"domain_scores_codex":[0.9966589,0.00002980467,0.00122815,0.001193039,0.00007827977,0.0008118364],"domain_scores_gemma":[0.997597,0.0001302876,0.0009300435,0.001034748,0.0001732406,0.0001347272],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"not_applicable","study_design_scores_codex":[0.000206534,0.0002511488,0.9424981,0.001425722,0.0004384143,0.00003531891,0.0007328175,0.01929756,0.000005494142,0.004973298,0.02698103,0.003154496],"study_design_scores_gemma":[0.002695204,0.0002540797,0.2041646,0.0003887217,0.00008513904,0.000002269358,0.0002843032,0.05525525,0.0001244076,0.004406504,0.7309757,0.001363785],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9164962,0.005581746,0.05568008,0.00235693,0.004367497,0.002777728,0.0005587583,0.0003598804,0.01182115],"genre_scores_gemma":[0.8470844,0.002216725,0.001573502,0.0005671243,0.0005583692,0.0008663681,0.0006257964,0.00008862535,0.1464191],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7383336,"threshold_uncertainty_score":0.9995793,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06330541523838851,"score_gpt":0.2633495519107036,"score_spread":0.2000441366723151,"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."}}