{"id":"W2922444559","doi":"10.24963/ijcai.2019/66","title":"Computing Approximate Equilibria in Sequential Adversarial Games by Exploitability Descent","year":2019,"lang":"en","type":"preprint","venue":"","topic":"Advanced Bandit Algorithms Research","field":"Decision Sciences","cited_by":18,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"Alberta Machine Intelligence Institute","keywords":"Fictitious play; Convergence (economics); Nash equilibrium; Computer science; Counterfactual thinking; Benchmark (surveying); Mathematical optimization; Perfect information; Regret; Descent (aeronautics); Class (philosophy); Zero (linguistics); Imperfect; Mathematical economics; Mathematics; Economics; Artificial intelligence","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":["metaepi_narrow","open_science","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.009937959,0.000582198,0.001280035,0.0006362904,0.0001314555,0.0008730745,0.003114317,0.0005880599,0.001663487],"category_scores_gemma":[0.005465235,0.0004673949,0.0004007294,0.0007949951,0.0003705536,0.0005664912,0.008073648,0.002224215,0.0006746968],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007631243,"about_ca_system_score_gemma":0.0007399292,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005340254,"about_ca_topic_score_gemma":0.0000940434,"domain_scores_codex":[0.9886282,0.001464341,0.002000412,0.00263505,0.004086873,0.001185139],"domain_scores_gemma":[0.9933455,0.00277282,0.0006121352,0.002409184,0.0005728174,0.0002875642],"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":[0.002693057,0.002904112,0.07827062,0.0009572558,0.0003823143,0.0002412432,0.008029146,0.4683293,0.01189653,0.0008121552,0.08301345,0.3424708],"study_design_scores_gemma":[0.002429199,0.00009778253,0.002613502,0.00016543,0.00001351106,0.000006524717,0.001718391,0.9344377,0.003896782,0.04940857,0.00413364,0.001078985],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3923278,0.0002959724,0.5979472,0.0004444221,0.003792992,0.002104257,0.0001923209,0.0001846129,0.002710452],"genre_scores_gemma":[0.9827365,0.00003148386,0.01435838,0.00006029532,0.0004267927,0.00004400103,0.0001124635,0.00005765481,0.00217245],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5904087,"threshold_uncertainty_score":0.9999489,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.156201570884926,"score_gpt":0.4414865312104221,"score_spread":0.2852849603254962,"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."}}