{"id":"W1512866498","doi":"10.1609/aaai.v26i1.8321","title":"Investigating Contingency Awareness Using Atari 2600 Games","year":2021,"lang":"en","type":"article","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":76,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"University of Alberta; Alberta Innovates; Western Canada Research Grid; Compute Canada","keywords":"Contingency; Reinforcement learning; Exploit; Computer science; Function (biology); Control (management); Value (mathematics); Reinforcement; Artificial intelligence; Contingency management; Bellman equation; Human–computer interaction; Machine learning; Psychology; Social psychology; Computer security","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001858963,0.0007595566,0.0006256036,0.0004779931,0.0004931751,0.0009255846,0.001260298,0.0006194917,0.001627991],"category_scores_gemma":[0.01144448,0.0002935231,0.0004782608,0.0002868284,0.001110954,0.001582765,0.001276066,0.001677876,0.0001290029],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001187535,"about_ca_system_score_gemma":0.0006866451,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008297101,"about_ca_topic_score_gemma":0.009641239,"domain_scores_codex":[0.9989699,0.0005701605,0.00003224386,0.000163414,0.0001790777,0.0000851654],"domain_scores_gemma":[0.9949379,0.003808554,0.000361906,0.0003247825,0.0002736406,0.0002932895],"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":[0.001081188,0.001101043,0.01734096,0.000248173,0.0002520932,0.0004110586,0.001276351,0.879957,0.007775597,0.03053759,0.002213333,0.05780557],"study_design_scores_gemma":[0.00004610807,0.0002821029,0.002141854,0.000009630457,0.00001213899,0.00003159877,0.0001453973,0.9865964,0.001304879,0.008717217,0.000695378,0.00001727218],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9327192,0.000123397,0.06183095,0.0002207875,0.00002421628,0.0001456202,0.0001804084,0.0002237464,0.004531718],"genre_scores_gemma":[0.978865,0.00002879425,0.02011225,0.00003834449,0.000004557876,0.00005627756,0.0001340489,0.00001761086,0.0007431327],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.008297101,"threshold_uncertainty_score":0.01649761,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1387859785918386,"score_gpt":0.3272622020942906,"score_spread":0.188476223502452,"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."}}