{"id":"W6928694211","doi":"10.48448/rf5w-dr12","title":"Lexical Substitution as Causal Language Modeling","year":2024,"lang":"en","type":"other","venue":"Open MIND","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Language model; Substitution (logic); Causal model; Language understanding; Autoregressive model; Task (project management); Causal analysis","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":[0.001780059,0.0009355173,0.0006284069,0.001269172,0.0006048541,0.002321813,0.001664677,0.001124815,0.01418302],"category_scores_gemma":[0.01034663,0.0005825168,0.001055227,0.001223463,0.001195322,0.004320466,0.00244181,0.001855583,0.005788561],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000760008,"about_ca_system_score_gemma":0.001747728,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002289675,"about_ca_topic_score_gemma":0.004544435,"domain_scores_codex":[0.9980247,0.00101404,0.00009909776,0.0004743757,0.0002969135,0.00009085175],"domain_scores_gemma":[0.9966477,0.002065009,0.000198381,0.0007127812,0.000288817,0.00008727086],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0003334362,0.0001869871,0.004689267,0.0007019247,0.0001505038,0.0007649931,0.001315154,0.08220933,0.01291688,0.2556575,0.03865382,0.6024203],"study_design_scores_gemma":[0.00002995021,0.00004498856,0.0004556567,0.00006279209,0.00003563602,0.0003024612,0.0001971751,0.7438273,0.00863281,0.2149975,0.03137261,0.00004112939],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01069006,0.0002839003,0.9647606,0.0007258123,0.0001656731,0.00009781399,0.001265931,0.01382238,0.008187792],"genre_scores_gemma":[0.4544331,0.0004181571,0.5263981,0.00067522,0.0001825153,0.0002774416,0.004308507,0.002513481,0.01079351],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01418302,"threshold_uncertainty_score":0.04744691,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06420371864228186,"score_gpt":0.3782639194142832,"score_spread":0.3140602007720014,"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."}}