{"id":"W4404781255","doi":"10.18653/v1/2024.findings-emnlp.790","title":"Topic Modeling: Contextual Token Embeddings Are All You Need","year":2024,"lang":"en","type":"article","venue":"","topic":"Computational and Text Analysis Methods","field":"Social Sciences","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa","funders":"","keywords":"Security token; Computer science; Computer security","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":[0.002805914,0.001473157,0.001052819,0.001562614,0.0005948539,0.002736842,0.001192637,0.00121681,0.005562848],"category_scores_gemma":[0.0154388,0.0005465559,0.001004278,0.002386847,0.0006362658,0.01045947,0.001574325,0.002573909,0.004316085],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006450508,"about_ca_system_score_gemma":0.001192588,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003009578,"about_ca_topic_score_gemma":0.004712314,"domain_scores_codex":[0.9979875,0.0008773176,0.000151623,0.0004788439,0.0003859044,0.000118863],"domain_scores_gemma":[0.9946596,0.002531682,0.0004459519,0.001235085,0.0009139711,0.0002136674],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0008057338,0.0003953546,0.01612406,0.0009757901,0.0005662109,0.0002390057,0.001075555,0.06933756,0.008787222,0.07294331,0.0970658,0.7316843],"study_design_scores_gemma":[0.00008343439,0.0002457152,0.003720322,0.0003195306,0.0002771379,0.0005330833,0.0007521236,0.6911792,0.009911467,0.2099036,0.08292489,0.0001494331],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01847741,0.002734725,0.9614528,0.003068849,0.0006462593,0.0002074847,0.002935347,0.005696011,0.004781032],"genre_scores_gemma":[0.4054269,0.004045819,0.5671899,0.001381194,0.001132843,0.0006092191,0.01098222,0.001906962,0.00732488],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005562848,"threshold_uncertainty_score":0.01860958,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09091584989938681,"score_gpt":0.4075066138841508,"score_spread":0.3165907639847639,"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."}}