{"id":"W4389524498","doi":"10.18653/v1/2023.emnlp-industry.26","title":"Unveiling Identity Biases in Toxicity Detection : A Game-Focused Dataset and Reactivity Analysis Approach","year":2023,"lang":"en","type":"article","venue":"","topic":"Hate Speech and Cyberbullying Detection","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Ubisoft (Canada)","funders":"Mitacs","keywords":"Computer science; Conversation; Identity (music); Harm; Class (philosophy); Benchmarking; Perspective (graphical); Artificial intelligence; Natural language processing; Machine learning; Psychology; Social psychology; Communication","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":[0.0009432144,0.0001109916,0.0001862248,0.0006914658,0.0001082977,0.0002424183,0.0002493918,0.00006780795,0.000008609039],"category_scores_gemma":[0.0001782345,0.0001059332,0.00005422538,0.004463645,0.00002556053,0.001256644,0.0002223061,0.0001462934,0.000040323],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000517771,"about_ca_system_score_gemma":0.00001960188,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001928812,"about_ca_topic_score_gemma":0.004421502,"domain_scores_codex":[0.9986691,0.0001265968,0.0001830109,0.0005079026,0.0002401072,0.0002732468],"domain_scores_gemma":[0.9992772,0.0001173703,0.00005889263,0.0004397035,0.00002236257,0.00008452062],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001387987,0.0009672425,0.04359725,0.0001581539,0.0006961082,0.0002041261,0.002108719,0.02635542,0.1842893,0.001844889,0.001182615,0.7384574],"study_design_scores_gemma":[0.0002502806,0.00003562977,0.08567141,0.000005363135,0.0000378533,0.000007974431,0.00006936643,0.8822079,0.03088435,0.0004156593,0.0002295775,0.000184687],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6584311,0.000008289563,0.340964,0.00003794807,0.00005101288,0.00008487848,0.00001753045,0.0002371618,0.0001680237],"genre_scores_gemma":[0.9972282,0.00002386826,0.002537359,0.0000367896,0.00001876352,0.0000146766,0.00007321194,0.000004278591,0.00006285886],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8558524,"threshold_uncertainty_score":0.4319829,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03807695971279029,"score_gpt":0.2792286478311847,"score_spread":0.2411516881183944,"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."}}