{"id":"W4415089370","doi":"10.1080/02699931.2025.2568551","title":"Affective connotations according to LLMs: implications for meaning measurement and cultural bias","year":2025,"lang":"en","type":"article","venue":"Cognition & Emotion","topic":"Computational and Text Analysis Methods","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Agence Nationale de la Recherche; Deutsche Forschungsgemeinschaft","keywords":"Meaning (existential); Scholarship; German; Generative grammar; Dimension (graph theory); Connotation; Social influence; Test (biology)","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":[],"consensus_categories":[],"category_scores_codex":[0.0009294716,0.00006337735,0.00009392316,0.0001634642,0.0009470847,0.0001238818,0.00004703266,0.00003502189,0.00001180921],"category_scores_gemma":[0.001778094,0.00006508602,0.00005848901,0.0005394137,0.00004955122,0.0001936232,0.00001771759,0.00003794118,0.000006084077],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001346789,"about_ca_system_score_gemma":0.00008027641,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008447163,"about_ca_topic_score_gemma":0.0004476201,"domain_scores_codex":[0.9992117,0.0001881001,0.0001465942,0.0001892735,0.0001465365,0.0001177616],"domain_scores_gemma":[0.9983933,0.0004151748,0.00006073351,0.00004392549,0.001030627,0.00005623681],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00002321815,0.00007532851,0.0034124,0.00002700037,0.0001185452,6.772134e-8,0.008982577,0.000260932,0.00572541,0.5211271,0.0009224751,0.459325],"study_design_scores_gemma":[0.001280767,0.0001039319,0.2518537,0.0003524049,0.0006110467,0.000001108101,0.04099694,0.002724333,0.001369882,0.6840398,0.01622816,0.0004378973],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09855092,0.00005932041,0.8704224,0.01005411,0.000218562,0.0007812387,0.00001822833,0.00009488253,0.01980033],"genre_scores_gemma":[0.9916304,0.00001379208,0.007540642,0.0003507232,0.0001035051,0.0001559508,0.00004367091,0.000003359246,0.0001579266],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8930795,"threshold_uncertainty_score":0.7284305,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2070737795253065,"score_gpt":0.4406527657466264,"score_spread":0.2335789862213199,"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."}}