{"id":"W4390722673","doi":"10.48550/arxiv.2401.02974","title":"Efficacy of Utilizing Large Language Models to Detect Public Threat Posted Online","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Hate Speech and Cyberbullying Detection","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Centennial College","funders":"California State University Long Beach","keywords":"Moderation; Transparency (behavior); Identification (biology); Scale (ratio); Computer security; Internet privacy; Advertising; Computer science; Psychology; Business; Social psychology; Geography","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.03013776,0.002043826,0.001118161,0.00319597,0.0008655714,0.003247295,0.001164708,0.001484436,0.001996109],"category_scores_gemma":[0.1140092,0.0006237859,0.001216786,0.001001676,0.0008303749,0.005962016,0.002370474,0.002359106,0.00201825],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000938058,"about_ca_system_score_gemma":0.001500644,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004024488,"about_ca_topic_score_gemma":0.0060735,"domain_scores_codex":[0.9779545,0.01750032,0.0006731495,0.001955332,0.001549144,0.0003676011],"domain_scores_gemma":[0.8196026,0.1643478,0.004804063,0.005917805,0.00398191,0.001345876],"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":[0.005013342,0.003457427,0.2990257,0.001006262,0.001799119,0.0003806804,0.0039907,0.08543344,0.01218615,0.003977804,0.0121701,0.5715593],"study_design_scores_gemma":[0.0001789169,0.001445206,0.03148351,0.0001567021,0.0005430516,0.0002513078,0.001521148,0.9418034,0.00926084,0.008911688,0.004249211,0.0001950394],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8222587,0.001157303,0.1520099,0.003498425,0.0003191635,0.0009253339,0.0022428,0.005466611,0.01212175],"genre_scores_gemma":[0.9453281,0.000184929,0.05101296,0.0003854421,0.0001187163,0.000225247,0.001398741,0.0001221599,0.001223697],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.03013776,"threshold_uncertainty_score":0.1593856,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1278831013225399,"score_gpt":0.2267668160144624,"score_spread":0.09888371469192248,"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."}}