{"id":"W4291821302","doi":"10.32470/ccn.2022.1229-0","title":"Continual Reinforcement Learning with Multi-Timescale Successor Features","year":2022,"lang":"en","type":"article","venue":"2022 Conference on Cognitive Computational Neuroscience","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":false,"ca_institutions":"McGill University","funders":"","keywords":"Successor cardinal; Reinforcement learning; Reinforcement; Computer science; Artificial intelligence; Engineering; Structural engineering; Mathematics","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","sts"],"consensus_categories":[],"category_scores_codex":[0.0004727086,0.0003141097,0.0002489521,0.0003088769,0.001560275,0.0005008864,0.001391929,0.0000305893,0.0002355112],"category_scores_gemma":[0.0003240948,0.0003015963,0.00007079392,0.00113097,0.0003162814,0.0006133451,0.001002928,0.0009261509,0.00004975396],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001081972,"about_ca_system_score_gemma":0.0004639198,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001244956,"about_ca_topic_score_gemma":0.000001915745,"domain_scores_codex":[0.9958309,0.0003881073,0.0003467231,0.000995059,0.00188236,0.0005568629],"domain_scores_gemma":[0.998201,0.0004993026,0.0003591493,0.0002689478,0.0004955978,0.0001759992],"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":[0.00006603684,0.00009633431,0.0008794645,0.00000668019,0.000008604165,0.00007264468,0.000500937,0.964101,0.0003488774,0.03136614,0.0001559901,0.002397355],"study_design_scores_gemma":[0.001242642,0.00176399,0.01273447,0.00004020872,0.000009023957,0.00008289801,0.0003583936,0.98205,0.0002559176,0.0002254816,0.0008181431,0.0004188319],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.008372466,0.00001499231,0.9846889,0.0008527588,0.0004699548,0.0005150656,0.00001156129,0.0002586415,0.004815594],"genre_scores_gemma":[0.9862511,0.000007139386,0.004809146,0.002640358,0.00002421295,0.0001500184,0.00005188357,0.00002008813,0.006046088],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9798798,"threshold_uncertainty_score":0.9999436,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0353669370020116,"score_gpt":0.2839279176125578,"score_spread":0.2485609806105462,"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."}}