{"id":"W4414602263","doi":"10.1057/s41599-025-05834-4","title":"LLMs as annotators: the effect of party cues on labelling decisions by large language models","year":2025,"lang":"en","type":"article","venue":"Humanities and Social Sciences Communications","topic":"Topic Modeling","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Western University","funders":"Universität Wien","keywords":"Politics; Statement (logic); Test (biology); Motivated reasoning; Replicate","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":["sts"],"consensus_categories":[],"category_scores_codex":[0.0007747162,0.00007312135,0.0001179131,0.00005838862,0.002378372,0.0002117656,0.001695567,0.00003124076,0.000003768205],"category_scores_gemma":[0.00005388573,0.00004955545,0.00004174549,0.0002521237,0.0004992247,0.0002465151,0.0005917794,0.0001243184,0.000001707858],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001486401,"about_ca_system_score_gemma":0.00004067234,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003180375,"about_ca_topic_score_gemma":0.0002402953,"domain_scores_codex":[0.9991097,0.000229642,0.0001632878,0.0001505885,0.0001893491,0.0001574329],"domain_scores_gemma":[0.9982876,0.001039218,0.00006624214,0.0005592292,0.00003702235,0.000010755],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[9.144085e-7,0.00001875981,0.0001113978,0.00000396433,0.000006367801,6.622035e-8,0.01304672,0.0000736167,0.00002825064,0.9810466,0.0007515596,0.00491178],"study_design_scores_gemma":[0.001064703,0.0005789624,0.0005033031,0.0003468499,0.00006890763,0.000001920692,0.0403014,0.6215864,0.001339578,0.2559609,0.07766937,0.0005777464],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.742083,0.008690531,0.1183567,0.01231466,0.0003115041,0.000592997,0.00004814267,0.0001894949,0.1174131],"genre_scores_gemma":[0.9974443,0.0002656067,0.001151066,0.0007171978,0.00001235235,0.00002268404,0.000002771553,0.000001933413,0.0003821242],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7250857,"threshold_uncertainty_score":0.9989204,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07625001237173913,"score_gpt":0.3429990491010771,"score_spread":0.266749036729338,"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."}}