{"id":"W4412496371","doi":"10.1016/j.jml.2025.104663","title":"Learning filler-gap dependencies with neural language models: Testing island sensitivity in Norwegian and English","year":2025,"lang":"en","type":"article","venue":"Journal of Memory and Language","topic":"Phonetics and Phonology Research","field":"Psychology","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"Social Sciences and Humanities Research Council; Social Sciences and Humanities Research Council of Canada","keywords":"Norwegian; Psychology; Filler (materials); Linguistics; Cognitive psychology; Sensitivity (control systems)","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":[0.0008479859,0.0001019963,0.0002340418,0.0002128687,0.00007954731,0.00003900731,0.00006735433,0.00008778802,0.00005064729],"category_scores_gemma":[0.0002497382,0.00007762574,0.00002198026,0.0001688338,0.00009945069,0.0001061627,0.00005724371,0.0007306182,6.345094e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001213709,"about_ca_system_score_gemma":0.00003946303,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003456249,"about_ca_topic_score_gemma":0.001080522,"domain_scores_codex":[0.9990798,0.0002595163,0.0001897027,0.0001432137,0.000110296,0.0002174448],"domain_scores_gemma":[0.9991859,0.0004735058,0.0001005756,0.00009539021,0.0000795283,0.00006515367],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"observational","study_design_scores_codex":[0.001839053,0.000325782,0.2608792,0.0002592852,0.0003751381,0.01442712,0.5469984,0.003251952,0.03520903,0.0002989835,0.0007764408,0.1353596],"study_design_scores_gemma":[0.007518306,0.001609179,0.6804266,0.0005736535,0.0001187153,0.003340019,0.2931044,0.01030352,0.001888324,0.0004570636,0.0001366823,0.0005235047],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9841354,0.002604636,0.00004997558,0.0001243428,0.0001199673,0.00006284688,0.000003741513,0.00000918561,0.01288983],"genre_scores_gemma":[0.9984734,0.0000329527,0.0001591514,0.0001089015,0.000100281,0.000001358679,0.00000102479,0.000007879766,0.00111502],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4195474,"threshold_uncertainty_score":0.3174213,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02322602444790397,"score_gpt":0.3009518162577645,"score_spread":0.2777257918098605,"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."}}