{"id":"W4405618562","doi":"10.1111/emip.12663","title":"Instruction‐Tuned Large‐Language Models for Quality Control in Automatic Item Generation: A Feasibility Study","year":2024,"lang":"en","type":"article","venue":"Educational Measurement Issues and Practice","topic":"Topic Modeling","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Quality (philosophy); Computer science; Control (management); Item response theory; Language proficiency; Mathematics education; Natural language processing; Psychology; Artificial intelligence; Psychometrics; Developmental psychology","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":[0.03405164,0.001157849,0.0007550939,0.000753626,0.000533775,0.001884429,0.003070298,0.001326082,0.00270793],"category_scores_gemma":[0.1322777,0.000945216,0.0005778416,0.0007102212,0.000917786,0.00323792,0.001626935,0.002113918,0.0008768544],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001603922,"about_ca_system_score_gemma":0.001880966,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008011228,"about_ca_topic_score_gemma":0.005956651,"domain_scores_codex":[0.9849164,0.01145417,0.0009187471,0.00113738,0.001252314,0.0003209713],"domain_scores_gemma":[0.8163837,0.1545434,0.003561013,0.01226614,0.01161165,0.001634093],"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":[0.01147016,0.02063601,0.08636843,0.001161271,0.0004352772,0.0009060684,0.007445504,0.1733389,0.06225172,0.005511854,0.0082977,0.6221771],"study_design_scores_gemma":[0.001451004,0.004709852,0.0165485,0.00009154419,0.000152971,0.0001976538,0.0005376074,0.9450889,0.02638712,0.002219969,0.002486016,0.0001288998],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8504725,0.0001305987,0.1416808,0.000483351,0.00006182345,0.002063565,0.0002938132,0.003482544,0.001330942],"genre_scores_gemma":[0.8636357,0.00002986427,0.1345947,0.000114437,0.0000154556,0.0008314723,0.0002336638,0.0002096155,0.0003351135],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.03405164,"threshold_uncertainty_score":0.1800845,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1967005621322491,"score_gpt":0.4278503410420334,"score_spread":0.2311497789097843,"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."}}