{"id":"W6902733987","doi":"10.7910/dvn/yaulad","title":"Data for \"Iterative LLM-Guided Sampling and Expert-annotated Benchmark Corpus for Harmful Suicide Content Detection\"","year":2024,"lang":"en","type":"dataset","venue":"Harvard Dataverse","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kootenay Association for Science & Technology","funders":"","keywords":"Benchmark (surveying); Harm; Leverage (statistics); The Internet; Task (project management); Poison control; Suicide prevention; Human factors and ergonomics","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":[0.003524182,0.002323083,0.0008839125,0.003951856,0.001634327,0.001664704,0.003852471,0.003330997,0.02389051],"category_scores_gemma":[0.01420648,0.0005782532,0.001144149,0.00430895,0.001018118,0.001667357,0.002925599,0.002508881,0.03473974],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00243278,"about_ca_system_score_gemma":0.002557527,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.04761586,"about_ca_topic_score_gemma":0.1285561,"domain_scores_codex":[0.9969903,0.0009708718,0.0003599607,0.0006715551,0.0007180964,0.0002891737],"domain_scores_gemma":[0.9919337,0.002421722,0.0003957512,0.001970444,0.002779544,0.0004988383],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0001445709,0.0001310016,0.002367973,0.0005219064,0.00003791708,0.0001216391,0.0001079993,0.0006858538,0.0004451939,0.0006528813,0.9847738,0.01000937],"study_design_scores_gemma":[0.000500947,0.00009817523,0.01604497,0.0003680943,0.00006508919,0.0004227003,0.0007525705,0.006848313,0.002851506,0.002505273,0.9694286,0.0001138128],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.005184807,0.0003799559,0.001306237,0.0008105987,0.0002833838,0.0001883988,0.9856862,0.002374431,0.003785888],"genre_scores_gemma":[0.003182052,0.00005140125,0.002143633,0.0001499147,0.00001735407,0.0002308962,0.9926229,0.0001093211,0.001492571],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.04761586,"threshold_uncertainty_score":0.09467739,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1745758825352345,"score_gpt":0.3598631031370935,"score_spread":0.1852872206018589,"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."}}