{"id":"W4413049767","doi":"10.1016/j.jfds.2025.100165","title":"Catastrophic-risk-aware reinforcement learning with extreme-value-theory-based policy gradients☆","year":2025,"lang":"en","type":"article","venue":"The Journal of Finance and Data Science","topic":"Risk and Portfolio Optimization","field":"Decision Sciences","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Reinforcement learning; Extreme value theory; Reinforcement; Computer science; Risk analysis (engineering); Business; Artificial intelligence; Mathematics; Psychology; Social psychology; Statistics","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":[0.009989196,0.0001090221,0.0001917021,0.0005044133,0.0008504112,0.0002739952,0.00235044,0.000022182,0.00001204881],"category_scores_gemma":[0.002368415,0.00005441098,0.00002441628,0.002412552,0.000769506,0.001486368,0.0003209719,0.0002577433,0.00000559948],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003672007,"about_ca_system_score_gemma":0.0008895032,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00009688802,"about_ca_topic_score_gemma":0.00001429435,"domain_scores_codex":[0.9975861,0.0002084187,0.0005127367,0.0002751683,0.00116289,0.000254721],"domain_scores_gemma":[0.9972265,0.0004891892,0.0008301476,0.0009211466,0.0004579449,0.00007506422],"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":[0.0006186482,0.00004545265,0.02892606,0.000004629909,0.00001581722,0.00001350947,0.0008193579,0.7268498,0.0002255628,0.02246207,0.002604533,0.2174146],"study_design_scores_gemma":[0.003447097,0.001808165,0.1729588,0.000627902,0.00025893,0.0002329529,0.003221483,0.6476005,0.002249183,0.03082426,0.1361618,0.0006089089],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3791559,0.0006614051,0.6159969,0.00258375,0.0002297153,0.0001540738,0.0000340929,0.000008700766,0.00117548],"genre_scores_gemma":[0.9935565,0.002917862,0.002529868,0.0002747115,0.00005030878,5.553601e-7,0.000003358237,0.000003176385,0.0006636517],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6144006,"threshold_uncertainty_score":0.6540762,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05512067823412605,"score_gpt":0.3696084126283005,"score_spread":0.3144877343941745,"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."}}