{"id":"W4416193684","doi":"10.1080/23273798.2025.2585303","title":"Surprisal in reading: language models predict the N400 for L2 readers","year":2025,"lang":"en","type":"article","venue":"Language Cognition and Neuroscience","topic":"Text Readability and Simplification","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"Graduate Student Association, Fordham University; National Science Foundation of Sri Lanka; Alliance de recherche numérique du Canada; Social Sciences and Humanities Research Council of Canada; Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research","keywords":"N400; Language model; Semantics (computer science); Computational linguistics; Utterance","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0008565832,0.0004332797,0.0002150515,0.0003854212,0.000181818,0.001188556,0.0001692067,0.0005584485,0.001913995],"category_scores_gemma":[0.009985163,0.0002321251,0.0002559212,0.0001704869,0.000336407,0.000993353,0.0005786524,0.0005661726,0.0003754409],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002008658,"about_ca_system_score_gemma":0.0001806446,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002024055,"about_ca_topic_score_gemma":0.002415858,"domain_scores_codex":[0.9997286,0.00008553463,0.00002136211,0.00007616965,0.00006330707,0.00002498728],"domain_scores_gemma":[0.9949247,0.003597089,0.0007075919,0.0003089532,0.000302719,0.0001589666],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00247623,0.000532236,0.6607308,0.0003270317,0.00028371,0.002291839,0.0186576,0.004010235,0.2612842,0.001214621,0.0006444516,0.04754705],"study_design_scores_gemma":[0.00003606476,0.0004421738,0.9618382,0.00003311538,0.00009016867,0.001235661,0.002623009,0.01710715,0.01373066,0.002349861,0.0004545241,0.00005948428],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9977366,0.00002561976,0.001116791,0.0000320247,0.000003420819,0.000009274218,0.0000318261,0.00003129174,0.001013151],"genre_scores_gemma":[0.999378,0.00001379072,0.0003326291,0.00001034573,0.00000266472,0.000005174308,0.00003164133,0.00001204151,0.0002137735],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.002024055,"threshold_uncertainty_score":0.006402969,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02908256521657047,"score_gpt":0.2939134694958276,"score_spread":0.2648309042792572,"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."}}