{"id":"W4411119799","doi":"10.18653/v1/2025.naacl-long.118","title":"Fine-Tuned LLMs are “Time Capsules” for Tracking Societal Bias Through Books","year":2025,"lang":"en","type":"article","venue":"","topic":"Hate Speech and Cyberbullying Detection","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Brock University","funders":"","keywords":"Tracking (education); Computer science; Political science; Sociology","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":[0.0008581488,0.0003964936,0.0004592355,0.002842508,0.0005070705,0.002141744,0.0007813828,0.0005695878,0.01731049],"category_scores_gemma":[0.01103604,0.0002726366,0.0002668893,0.002824284,0.0003257062,0.003371429,0.001479479,0.0006954438,0.00940384],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006132562,"about_ca_system_score_gemma":0.0003639794,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004188858,"about_ca_topic_score_gemma":0.01340273,"domain_scores_codex":[0.9995127,0.0001361688,0.00002559519,0.0001576935,0.0001235398,0.00004425706],"domain_scores_gemma":[0.9969796,0.001315825,0.0004175732,0.0005363945,0.0005879338,0.0001627863],"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":[0.0009814936,0.0001912024,0.08276939,0.0003385526,0.0001772405,0.000107637,0.001253331,0.004890233,0.007484623,0.01107652,0.08447336,0.8062564],"study_design_scores_gemma":[0.0001535489,0.0004254173,0.2420619,0.0002729435,0.0002381973,0.0003313997,0.003324857,0.4612978,0.01937923,0.06219095,0.2101066,0.000217123],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.545193,0.00726517,0.2560065,0.003923693,0.001353468,0.000421434,0.04085188,0.02979532,0.1151897],"genre_scores_gemma":[0.893676,0.0005091048,0.08043347,0.0003425251,0.0003023892,0.0001898464,0.008264379,0.001233103,0.0150492],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01731049,"threshold_uncertainty_score":0.05790937,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0365675797256866,"score_gpt":0.2754714269060195,"score_spread":0.2389038471803329,"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."}}