{"id":"W2966537673","doi":"10.24963/ijcai.2019/840","title":"LTL and Beyond: Formal Languages for Reward Function Specification in Reinforcement Learning","year":2019,"lang":"en","type":"article","venue":"","topic":"Formal Methods in Verification","field":"Computer Science","cited_by":152,"is_retracted":false,"has_abstract":true,"ca_institutions":"Centre for Social Innovation; Vector Institute; University of Toronto","funders":"Comisión Nacional de Investigación Científica y Tecnológica; Natural Sciences and Engineering Research Council of Canada; Microsoft Research","keywords":"Reinforcement learning; Computer science; Function (biology); Artificial intelligence; Automaton; Representation (politics); Temporal difference learning; Reinforcement; Machine learning; Psychology","routes":{"ca_aff":true,"ca_fund":true,"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":[0.0007327448,0.00006572113,0.00007690958,0.0001081638,0.00004964247,0.00006256942,0.0001557391,0.00004365216,0.00002853995],"category_scores_gemma":[0.00005488573,0.00006082184,0.0000154572,0.0001461694,0.000009485861,0.0009636391,0.00006949441,0.00007758154,0.00003521792],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004276213,"about_ca_system_score_gemma":0.00001198272,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001226521,"about_ca_topic_score_gemma":0.000003569796,"domain_scores_codex":[0.9992936,0.00003410135,0.0001826271,0.0002084731,0.0001167284,0.0001645227],"domain_scores_gemma":[0.9995766,0.00005334714,0.00007810973,0.000228641,0.00003806865,0.00002522995],"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":[0.00004970622,0.00001400967,0.003044641,0.00004578396,0.000004732171,1.527048e-7,0.001015369,0.002764466,0.006632528,0.7852069,0.0001112647,0.2011105],"study_design_scores_gemma":[0.001217251,0.0008740746,0.06276667,0.00002303815,0.000005844409,0.000006478745,0.001234811,0.8895011,0.01264856,0.003642057,0.02775849,0.0003216576],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03989252,0.0000320214,0.9404048,0.0001606019,0.0001958405,0.0003731361,1.207658e-7,0.00008197486,0.01885893],"genre_scores_gemma":[0.6900669,0.00001536424,0.3074144,0.0001281405,0.00002739558,0.00003855718,0.000007526254,0.000004617745,0.002297051],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8867366,"threshold_uncertainty_score":0.2480243,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01755740490172809,"score_gpt":0.2801044758634209,"score_spread":0.2625470709616928,"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."}}