{"id":"W4410552311","doi":"10.1016/j.jss.2025.112493","title":"Variational Prefix Tuning for diverse and accurate code summarization using pre-trained language models","year":2025,"lang":"en","type":"article","venue":"Journal of Systems and Software","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Toronto; University of New Brunswick","funders":"Natural Sciences and Engineering Research Council of Canada; Connaught Fund; University of Toronto","keywords":"Automatic summarization; Prefix; Code (set theory); Computer science; Natural language processing; Programming language; Algorithm; Artificial intelligence; Linguistics","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":[0.001305502,0.001317768,0.001406224,0.001316867,0.0008809285,0.001318989,0.001907835,0.001835038,0.004610502],"category_scores_gemma":[0.006056296,0.0007681011,0.001089785,0.001248249,0.0007100973,0.002989978,0.001622792,0.002957399,0.002444246],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009093903,"about_ca_system_score_gemma":0.002362501,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01005664,"about_ca_topic_score_gemma":0.02288879,"domain_scores_codex":[0.9990672,0.0002281961,0.0000569715,0.0003324322,0.0001805051,0.0001346491],"domain_scores_gemma":[0.9976936,0.001434279,0.00008551231,0.000266992,0.0004250187,0.00009455921],"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":[0.0006185037,0.0002665661,0.001406377,0.0003050395,0.0001843572,0.0002348483,0.0002657062,0.3221436,0.03174498,0.008652174,0.02058268,0.6135952],"study_design_scores_gemma":[0.00002117535,0.00003890262,0.0001416382,0.00000837698,0.00001879542,0.00002190855,0.00003296085,0.9906527,0.003365387,0.004528148,0.001159398,0.00001061522],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04434987,0.00128417,0.9387566,0.0004938564,0.0002885697,0.0001269263,0.0008391974,0.01136574,0.002495054],"genre_scores_gemma":[0.4499238,0.0005153604,0.5300462,0.0006844359,0.0002762805,0.0004061506,0.007664482,0.002452298,0.008031059],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01005664,"threshold_uncertainty_score":0.01999623,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02418294193712048,"score_gpt":0.2990349434856573,"score_spread":0.2748520015485368,"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."}}