{"id":"W2954811041","doi":"10.18653/v1/s19-2092","title":"UNBNLP at SemEval-2019 Task 5 and 6: Using Language Models to Detect Hate Speech and Offensive Language","year":2019,"lang":"en","type":"article","venue":"","topic":"Hate Speech and Cyberbullying Detection","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of New Brunswick","funders":"","keywords":"Offensive; Computer science; Language model; Language identification; Task (project management); Natural language processing; Artificial intelligence; SemEval; Character (mathematics); Language understanding; Word (group theory); Speech recognition; Linguistics; Natural language","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":[0.0002189528,0.000179777,0.0001992434,0.0001396925,0.0001111567,0.0001764986,0.0002098513,0.00008033689,0.00004054463],"category_scores_gemma":[0.00001441308,0.0001564642,0.00003499017,0.0002337584,0.00002253363,0.0003474272,0.0004692129,0.0001019235,0.0001991641],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005835709,"about_ca_system_score_gemma":0.00002029693,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001117751,"about_ca_topic_score_gemma":0.0002186913,"domain_scores_codex":[0.9987084,0.00005242783,0.0001498995,0.0005256123,0.000225807,0.0003378125],"domain_scores_gemma":[0.9992092,0.00004893855,0.00004963614,0.0004602413,0.00005284506,0.0001791921],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00004636838,0.00001438733,0.0002293695,0.00005603558,0.00004273807,0.0001430234,0.007382855,0.001154491,0.8479751,0.0007595188,0.0005140945,0.1416821],"study_design_scores_gemma":[0.001476679,0.0005172556,0.0009829083,0.0001258258,0.00003601105,0.0007640995,0.001109689,0.3927454,0.5985598,0.001442126,0.001210514,0.001029729],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9228078,0.0003959451,0.07383876,0.000136324,0.000190668,0.0003436149,0.00000500382,0.0001678586,0.002114017],"genre_scores_gemma":[0.9462153,0.00002389622,0.04827587,0.0007490449,0.00004609082,0.00000296189,0.00000189571,0.00001872349,0.004666265],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3915909,"threshold_uncertainty_score":0.6380423,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01165980983610699,"score_gpt":0.2339918116001839,"score_spread":0.2223320017640769,"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."}}