{"id":"W4411337211","doi":"10.1109/sp61157.2025.00191","title":"SV-TrustEval-C: Evaluating Structure and Semantic Reasoning in Large Language Models for Source Code Vulnerability Analysis","year":2025,"lang":"en","type":"article","venue":"","topic":"Software Reliability and Analysis Research","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Open Text (Canada); University of Ottawa","funders":"","keywords":"Computer science; Programming language; Natural language processing; Code (set theory); Source code; Artificial intelligence; Set (abstract data type)","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002878138,0.000154909,0.0004307385,0.000525344,0.0002276146,0.0002497438,0.000533508,0.0001050756,0.0000341397],"category_scores_gemma":[0.001261051,0.0001304509,0.000177613,0.002481049,0.00004720278,0.0003439448,0.0004157807,0.0002432524,7.188472e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000913707,"about_ca_system_score_gemma":0.0001262849,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004565158,"about_ca_topic_score_gemma":0.001627767,"domain_scores_codex":[0.9976922,0.0003467426,0.0003790305,0.0007567746,0.0003849055,0.0004402941],"domain_scores_gemma":[0.9978907,0.001035012,0.00006171408,0.0007536998,0.0001803497,0.00007859081],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0000682448,0.0002816088,0.395189,0.0006557669,0.001045159,0.000007414713,0.00688503,0.4251132,0.001155524,0.03361805,0.00009206608,0.1358889],"study_design_scores_gemma":[0.0004643658,0.00002166964,0.009458353,0.00002854885,0.0001172122,6.121796e-7,0.0003443358,0.9782015,0.0003688374,0.01084948,0.00001560565,0.0001294786],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.414579,0.000231919,0.5844781,0.0003176804,0.00001354683,0.0002028155,0.000008817528,0.00006519802,0.0001029139],"genre_scores_gemma":[0.9555115,0.000008636805,0.04374807,0.0001497692,0.00001172233,0.00002800124,0.00001105333,0.000005133122,0.0005261509],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5530883,"threshold_uncertainty_score":0.5319632,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02268102776703042,"score_gpt":0.3626329116289783,"score_spread":0.3399518838619479,"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."}}