{"id":"W4412041733","doi":"10.1007/978-3-031-97570-7_18","title":"AggTruth: Contextual Hallucination Detection Using Aggregated Attention Scores in LLMs","year":2025,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Functional Brain Connectivity Studies","field":"Neuroscience","cited_by":1,"is_retracted":false,"has_abstract":false,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Computer science; Artificial intelligence; Natural language processing","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":[0.0003871333,0.0007342177,0.0005249768,0.0008549152,0.0002051309,0.0008105796,0.0006674468,0.000456442,0.0104229],"category_scores_gemma":[0.00135984,0.0002022333,0.0002925385,0.000562397,0.0001524741,0.0005160231,0.001055283,0.0003570065,0.003143511],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001637443,"about_ca_system_score_gemma":0.0001999894,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00162869,"about_ca_topic_score_gemma":0.003738413,"domain_scores_codex":[0.9998196,0.0000343827,0.00001148368,0.00004311067,0.00006731752,0.00002418922],"domain_scores_gemma":[0.9997543,0.0001115613,0.00001953322,0.00003828995,0.00004604022,0.00003034388],"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":[0.0007892276,0.00008276064,0.004615133,0.0001550144,0.00008901965,0.0003118428,0.0000861846,0.005239014,0.03356452,0.001255538,0.0226074,0.9312044],"study_design_scores_gemma":[0.0002057701,0.0006718318,0.04495927,0.00008857526,0.0001601624,0.001732158,0.0002851663,0.7438605,0.167719,0.01251631,0.02763839,0.0001628686],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1825123,0.001775306,0.7239264,0.0003629116,0.0004690996,0.0003269718,0.006203202,0.07460244,0.00982139],"genre_scores_gemma":[0.5700906,0.000523505,0.4088259,0.0001727469,0.0002749196,0.0002529977,0.005818831,0.001718312,0.01232216],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.0104229,"threshold_uncertainty_score":0.03486812,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0382325667169461,"score_gpt":0.2701886435554073,"score_spread":0.2319560768384612,"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."}}