{"id":"W4411449748","doi":"10.1145/3715735","title":"Hallucination Detection in Large Language Models with Metamorphic Relations","year":2025,"lang":"en","type":"article","venue":"Proceedings of the ACM on software engineering.","topic":"Topic Modeling","field":"Computer Science","cited_by":15,"is_retracted":false,"has_abstract":true,"ca_institutions":"York University; University of Calgary","funders":"","keywords":"Recall; Computer science; Margin (machine learning); Cognitive psychology; Psychology; Artificial intelligence; Machine learning","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":[0.004440343,0.001629003,0.0009697395,0.001602469,0.0005243564,0.001914111,0.001774262,0.001527613,0.001041329],"category_scores_gemma":[0.02785358,0.000586184,0.001437133,0.0008889667,0.001061565,0.003592783,0.00309998,0.001948031,0.0009720192],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009038101,"about_ca_system_score_gemma":0.001050172,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004968088,"about_ca_topic_score_gemma":0.007165468,"domain_scores_codex":[0.9957956,0.001646061,0.0003460269,0.001034442,0.001001494,0.0001762769],"domain_scores_gemma":[0.9839357,0.01138654,0.001241786,0.002166385,0.001018391,0.000251211],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.001295752,0.0004617297,0.03873779,0.00141619,0.001056921,0.002796693,0.001949632,0.2775883,0.03925931,0.006633818,0.02174796,0.607056],"study_design_scores_gemma":[0.00004082602,0.0001039912,0.001659157,0.00003574269,0.00006977429,0.0004904126,0.0002246524,0.9760089,0.00959558,0.009499172,0.002234491,0.00003731053],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2107225,0.002920418,0.7291935,0.001382689,0.0001659639,0.0002710044,0.002028531,0.05079232,0.002523038],"genre_scores_gemma":[0.8225754,0.0004025807,0.170176,0.0009175448,0.00007748167,0.000142934,0.003407765,0.000859108,0.001441369],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004968088,"threshold_uncertainty_score":0.02348304,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.00800568433846788,"score_gpt":0.2089088547174961,"score_spread":0.2009031703790282,"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."}}