{"id":"W7125580424","doi":"10.1109/cascon66301.2025.00121","title":"Evaluating Toxicity Understanding of LLM Agents","year":2025,"lang":"","type":"article","venue":"","topic":"Hate Speech and Cyberbullying Detection","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Toxicity; Perception; Trustworthiness; Order (exchange); Action (physics); Risk assessment","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":[0.001352879,0.0002041934,0.0002949097,0.0003133888,0.0003844453,0.0002111018,0.0006489086,0.0001449188,0.0003875089],"category_scores_gemma":[0.0002334886,0.000208438,0.0001586887,0.001458188,0.0001003519,0.0003691645,0.0004375478,0.000215204,0.00004504453],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003664814,"about_ca_system_score_gemma":0.0003549551,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001277802,"about_ca_topic_score_gemma":0.00003189445,"domain_scores_codex":[0.9976369,0.0002208196,0.0006074203,0.000555225,0.0005502129,0.0004294078],"domain_scores_gemma":[0.9987618,0.0001551641,0.000218059,0.0006159169,0.0001552638,0.00009375276],"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":[0.00008932022,0.0006527567,0.001224612,0.0005221587,0.000354464,0.00001495549,0.002122161,0.002963483,0.05004086,0.4974921,0.003419177,0.441104],"study_design_scores_gemma":[0.0008526304,0.0006153652,0.0008747569,0.0005345683,0.00006832756,0.000004570464,0.0006719952,0.7961075,0.1645906,0.03503546,0.0003488087,0.0002954421],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09552184,0.0001289214,0.8231189,0.0006778204,0.001936628,0.0002703587,0.00000130311,0.00008412015,0.07826005],"genre_scores_gemma":[0.9775091,0.00004179363,0.01480073,0.000363272,0.00003844432,0.000003753696,4.843037e-7,0.000006305646,0.007236133],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8819872,"threshold_uncertainty_score":0.8499854,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1342825005152352,"score_gpt":0.3670419346009326,"score_spread":0.2327594340856974,"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."}}