{"id":"W2904676870","doi":"10.7554/elife.32548","title":"Offline replay supports planning in human reinforcement learning","year":2018,"lang":"en","type":"article","venue":"eLife","topic":"Memory and Neural Mechanisms","field":"Neuroscience","cited_by":154,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"National Institute of Mental Health; John Templeton Foundation","keywords":"Reinforcement learning; Task (project management); Computer science; Artificial intelligence; Machine learning; Human–computer interaction","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":[0.0003452348,0.0001032308,0.0001201813,0.00009314094,0.0002267769,0.00002545756,0.0001500455,0.00004318855,0.0007395324],"category_scores_gemma":[0.0004259534,0.00009564332,0.00002559046,0.0001686051,0.00004955289,0.00012988,0.00008336226,0.0002149496,0.000198214],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001937609,"about_ca_system_score_gemma":0.00001442269,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000004726646,"about_ca_topic_score_gemma":0.000003328672,"domain_scores_codex":[0.9987797,0.00008656135,0.0002576891,0.0003201781,0.0002690396,0.0002868441],"domain_scores_gemma":[0.999555,0.0000600989,0.00009067092,0.0002019012,0.00001606661,0.00007626133],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00002876479,0.00001544775,0.0009343313,0.000005746754,5.651979e-7,0.0001383549,0.000644966,0.0005433726,0.9959509,0.0009251853,0.0006114635,0.0002008818],"study_design_scores_gemma":[0.0002672653,0.0003517577,0.000386591,0.00003983048,0.000001478073,0.00002185576,0.00004623212,0.00153734,0.9918299,0.0001097058,0.005292145,0.000115921],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9847775,0.000004669591,0.0002804377,0.00008606682,0.0003345468,0.0001222399,3.61024e-7,0.00009185692,0.01430236],"genre_scores_gemma":[0.99243,0.000003117958,0.00005230064,0.001607469,0.0002334339,0.000008067901,0.000002773437,0.00001290669,0.005649894],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.008652462,"threshold_uncertainty_score":0.8097361,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1017343992395071,"score_gpt":0.3625170537246353,"score_spread":0.2607826544851282,"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."}}