{"id":"W4405077419","doi":"10.1075/scl.118.09mil","title":"Modeling fine-grained sociolinguistic variation","year":2024,"lang":"en","type":"book-chapter","venue":"Studies in corpus linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Variation (astronomy); Computer science; Linguistics; Astrophysics; Physics; Philosophy","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0007324419,0.0002898349,0.0002675234,0.001101648,0.000492536,0.001177925,0.0005789989,0.00032663,0.00233197],"category_scores_gemma":[0.004581069,0.0002172201,0.0002568989,0.001230131,0.0009451016,0.001010159,0.0006883708,0.0007475683,0.0003378739],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001902902,"about_ca_system_score_gemma":0.001189189,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.1558919,"about_ca_topic_score_gemma":0.2614704,"domain_scores_codex":[0.999673,0.0001295397,0.000008943383,0.00009521322,0.00005854431,0.00003484916],"domain_scores_gemma":[0.9977036,0.001599529,0.0001719426,0.0002308325,0.0002388912,0.00005513331],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002782689,0.0002337101,0.1538184,0.0005395514,0.0002590302,0.0009920141,0.007373995,0.5049837,0.04215569,0.1118931,0.01190018,0.1655723],"study_design_scores_gemma":[0.00000677169,0.00002223223,0.05026158,0.0000272581,0.00001280529,0.00004729866,0.0007810646,0.916795,0.001619891,0.02372525,0.006673152,0.00002776742],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.7573251,0.0004555604,0.2290324,0.0005382867,0.00004509842,0.0000856229,0.00371875,0.000646351,0.008152733],"genre_scores_gemma":[0.9693927,0.00008760555,0.02744446,0.00004546264,0.00001007352,0.00005831815,0.001587796,0.00009273202,0.001280749],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.1558919,"threshold_uncertainty_score":0.309969,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05598935463493161,"score_gpt":0.3361349588186589,"score_spread":0.2801456041837273,"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."}}