{"id":"W4411119063","doi":"10.18653/v1/2025.nlp4dh-1.25","title":"Evaluating Large Language Models for Narrative Topic Labeling","year":2025,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"Social Sciences and Humanities Research Council of Canada","keywords":"Computer science; Narrative; Natural language processing; Linguistics; Philosophy","routes":{"ca_aff":true,"ca_fund":true,"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":[0.0005021827,0.00007811349,0.0001082614,0.00006457625,0.0001577128,0.00008329457,0.0004155794,0.00003886415,0.00001707978],"category_scores_gemma":[0.00009605721,0.00006947372,0.00004512626,0.0001598009,0.000004857524,0.0002934911,0.0001909666,0.0000556723,0.000002621074],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003379505,"about_ca_system_score_gemma":0.00007881153,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002343903,"about_ca_topic_score_gemma":0.00004231633,"domain_scores_codex":[0.9991479,0.00003286579,0.0001717792,0.0002959464,0.0001238111,0.0002276796],"domain_scores_gemma":[0.9993648,0.0001278751,0.00003096183,0.0003589869,0.00009221391,0.00002514084],"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":[0.000002108618,0.00002052511,0.00002398112,0.00002921396,0.00001677848,9.407488e-7,0.01126331,0.01387579,0.001368424,0.9031646,0.0003691991,0.06986511],"study_design_scores_gemma":[0.0003696971,0.00001704123,0.000003272934,0.00002340037,0.000003403,3.116654e-7,0.0007418876,0.9341562,0.001400333,0.06297527,0.0002377271,0.00007141662],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02695099,0.000212438,0.9621401,0.001153201,0.0002768117,0.0002435864,0.000001005852,0.0001635312,0.008858373],"genre_scores_gemma":[0.5591624,0.000001067824,0.4336371,0.001024666,0.00004175187,0.00004266907,0.00000142029,0.000003267983,0.0060856],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9202805,"threshold_uncertainty_score":0.2833056,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0665263816001478,"score_gpt":0.3842538718380301,"score_spread":0.3177274902378823,"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."}}