{"id":"W4393152329","doi":"10.1609/aaai.v38i21.30542","title":"Evaluating AI Red Teaming’s Readiness to Address Environmental Harms: A Thematic Analysis of LLM Discourse","year":2024,"lang":"en","type":"article","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","topic":"Law, AI, and Intellectual Property","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Thematic analysis; Thematic map; Psychology; Political science; Sociology; Geography; Qualitative research; Cartography; Social science","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":[0.001052192,0.0002852863,0.0005129346,0.0004578226,0.0001653144,0.0003746742,0.002275192,0.0000822506,0.0003253185],"category_scores_gemma":[0.0007081861,0.0001865058,0.0002969445,0.002044466,0.0004236804,0.000589919,0.0006048946,0.0002878873,0.00009209644],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00009299986,"about_ca_system_score_gemma":0.0001127625,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000573031,"about_ca_topic_score_gemma":0.00001114542,"domain_scores_codex":[0.9971296,0.00004235966,0.0008419807,0.00073468,0.0008707781,0.0003805899],"domain_scores_gemma":[0.9986145,0.0002155441,0.0002741214,0.0004897739,0.0002830782,0.0001230068],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001333006,0.000583199,0.0001497587,0.0002692858,0.0005866988,0.000001828058,0.02562206,0.000957368,0.4971561,0.2859412,0.0008824723,0.1877167],"study_design_scores_gemma":[0.00001521794,0.0003620484,0.00006460977,0.0005677786,0.0002173481,0.000002464702,0.001354057,0.5685933,0.4120581,0.01650648,0.0000361606,0.0002224736],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.839219,0.0001734017,0.1377117,0.007725368,0.001055603,0.001538709,0.00005179708,0.0002512043,0.01227327],"genre_scores_gemma":[0.9970721,0.00002755027,0.00184463,0.0002329835,0.00005289296,0.00005037449,0.000001430019,0.0000181906,0.0006998651],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5676359,"threshold_uncertainty_score":0.7605486,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1250722512946333,"score_gpt":0.3674303560178632,"score_spread":0.24235810472323,"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."}}