{"id":"W2998116579","doi":"10.1609/aaai.v34i04.6086","title":"Reinforcement Learning with Perturbed Rewards","year":2020,"lang":"en","type":"article","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":101,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"Defense Advanced Research Projects Agency; National Science Foundation","keywords":"Reinforcement learning; Computer science; Noise (video); Confusion matrix; Artificial intelligence; Convergence (economics); Confusion; Set (abstract data type); Gaussian; Matrix (chemical analysis); Mathematical optimization; Machine learning; Algorithm; 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":[],"consensus_categories":[],"category_scores_codex":[0.0003755465,0.0002544091,0.0002836983,0.00007214359,0.0002925585,0.0002496469,0.002232976,0.00007438039,0.00008798148],"category_scores_gemma":[0.0009574544,0.0001737393,0.00009541365,0.0008257778,0.0002168031,0.0005000687,0.0006382602,0.0006888905,0.00007546438],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004619964,"about_ca_system_score_gemma":0.0001210367,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002466723,"about_ca_topic_score_gemma":0.000001499386,"domain_scores_codex":[0.9978629,0.00002839361,0.0004552583,0.0005565387,0.0007350211,0.0003618235],"domain_scores_gemma":[0.9985555,0.00009009532,0.0004545815,0.0002395939,0.0005158082,0.000144374],"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":[0.000201974,0.00004143079,0.0005577088,0.00005918369,0.00003050019,0.000002074113,0.006326442,0.04833508,0.01269294,0.8967326,0.0001039337,0.03491617],"study_design_scores_gemma":[0.0000665672,0.0007882715,0.00009183104,0.0002081235,0.00001811935,0.000005546472,0.001092223,0.832746,0.1534239,0.0106284,0.0005876504,0.0003434149],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04464196,0.00001688809,0.8895583,0.02772583,0.0003301327,0.0006474149,6.381695e-7,0.0004134913,0.03666539],"genre_scores_gemma":[0.9898702,0.00001071058,0.00901443,0.0007051555,0.0001045763,0.00001519606,3.518026e-7,0.00001735094,0.0002620565],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9452282,"threshold_uncertainty_score":0.7084884,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0602824730324913,"score_gpt":0.2752996143925658,"score_spread":0.2150171413600745,"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."}}