{"id":"W3037571487","doi":"","title":"Value Preserving State-Action Abstractions","year":2020,"lang":"en","type":"article","venue":"International Conference on Artificial Intelligence and Statistics","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":18,"is_retracted":false,"has_abstract":false,"ca_institutions":"McGill University","funders":"","keywords":"Computer science; Value (mathematics); State (computer science); Action (physics); Programming language; Machine learning","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":[0.0001442417,0.0001386996,0.0001137312,0.00007962517,0.0001460135,0.0005362493,0.0005671955,0.00004300163,0.0003368287],"category_scores_gemma":[0.0004273273,0.0001433702,0.00002466306,0.0001183768,0.00007549416,0.0004266227,0.0001499639,0.000255421,0.0003660564],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000373681,"about_ca_system_score_gemma":0.00007477702,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005140958,"about_ca_topic_score_gemma":0.00001825552,"domain_scores_codex":[0.9986442,0.00005265069,0.0003645326,0.0003332616,0.0004204338,0.0001848655],"domain_scores_gemma":[0.9990258,0.0002501527,0.0001464699,0.000170865,0.0002636519,0.0001430561],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0000255792,0.00002141108,0.00003442695,0.000007740223,0.00001866919,0.00001222256,0.0007064152,0.1278758,0.0008141842,0.8014607,0.0003316262,0.06869115],"study_design_scores_gemma":[0.00002025178,0.0001673975,0.0002276003,0.00001584192,0.000003743577,0.000004462065,0.0002621328,0.9239734,0.002068251,0.07173079,0.001380387,0.0001457126],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0004589524,0.00000415901,0.9844529,0.005759501,0.0006266981,0.0001046169,0.00003407921,0.0001014268,0.008457677],"genre_scores_gemma":[0.9311973,0.0001494539,0.0673078,0.0008695347,0.0001198714,0.000008015429,0.0000222427,0.000009185019,0.000316607],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9307383,"threshold_uncertainty_score":0.5846468,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2378914496979285,"score_gpt":0.3741205699300081,"score_spread":0.1362291202320796,"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."}}