{"id":"W6910435347","doi":"10.48448/p9w5-af56","title":"Ask LLMs Directly, “What shapes your bias?”: Measuring Social Bias in Large Language Models","year":2024,"lang":"en","type":"other","venue":"Underline Science Inc.","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kootenay Association for Science & Technology","funders":"","keywords":"Viewpoints; Perception; Social perception; Social change; Social comparison theory; Social influence","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":[0.00797228,0.0007148855,0.000557181,0.00118857,0.0006617639,0.002361267,0.0005423193,0.0008975954,0.0018904],"category_scores_gemma":[0.05969794,0.0002114515,0.0004682736,0.0009611492,0.001156561,0.002643933,0.001859844,0.001192426,0.0007346159],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009176006,"about_ca_system_score_gemma":0.0007026068,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002841232,"about_ca_topic_score_gemma":0.003556858,"domain_scores_codex":[0.994348,0.003766214,0.0002816868,0.0005648085,0.0009035539,0.0001358055],"domain_scores_gemma":[0.9619212,0.02886468,0.003323711,0.003047356,0.002199261,0.0006437408],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.002101699,0.0005918958,0.4328894,0.001253086,0.001073989,0.0006297991,0.01290134,0.105678,0.02895533,0.04014866,0.01247043,0.3613063],"study_design_scores_gemma":[0.0001018826,0.0003975414,0.07861454,0.0002555173,0.0001880525,0.0003423265,0.003276421,0.7983745,0.01407475,0.09658795,0.007607412,0.0001791707],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8120068,0.001183488,0.1725762,0.002439233,0.0001402749,0.000202562,0.001986008,0.001143492,0.008321878],"genre_scores_gemma":[0.9731723,0.00011247,0.0249489,0.000225232,0.00005395823,0.0001005843,0.0008046108,0.0001162317,0.0004656406],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.00797228,"threshold_uncertainty_score":0.042162,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1319699398434422,"score_gpt":0.3462782630000419,"score_spread":0.2143083231565997,"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."}}