{"id":"W7131132546","doi":"10.1109/mlnlp66797.2025.11387796","title":"PersonaDrift: A Benchmark for Temporal Anomaly Detection in Language-Based Dementia Monitoring","year":2025,"lang":"","type":"article","venue":"","topic":"Persona Design and Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Benchmark (surveying); Anomaly detection; Baseline (sea); Statistical model; Hidden Markov model; Context (archaeology); Supervised learning; Replicate","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0004345893,0.0002344924,0.0002191387,0.0004030189,0.0003297042,0.0003026996,0.000598355,0.0001388794,0.00003752439],"category_scores_gemma":[0.00003722313,0.0002593459,0.0001573389,0.001312205,0.0000488841,0.0003073449,0.00008949331,0.0001752079,0.00001236183],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001861353,"about_ca_system_score_gemma":0.0003596114,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001277529,"about_ca_topic_score_gemma":0.000757657,"domain_scores_codex":[0.9981805,0.00006813979,0.0003887902,0.0006920092,0.0001932731,0.0004772611],"domain_scores_gemma":[0.9990904,0.0001585519,0.00009601748,0.0004755928,0.00008599898,0.00009348797],"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":[0.0001772108,0.001052061,0.09596548,0.0004467439,0.0002027262,0.00001936526,0.005063344,0.0003917663,0.1713831,0.01188332,0.001229597,0.7121853],"study_design_scores_gemma":[0.002204314,0.0001815397,0.04762952,0.0002191815,0.0000794083,0.000001623696,0.001258191,0.8124354,0.1309842,0.0004663587,0.004040467,0.0004997762],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1340835,0.001105431,0.8602021,0.001338659,0.0005361421,0.0008710421,0.000008260802,0.00009436095,0.001760511],"genre_scores_gemma":[0.9524473,0.000007380459,0.04533406,0.0002556176,0.0001081884,0.0004530821,0.000005211478,0.000009665446,0.001379517],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8183638,"threshold_uncertainty_score":0.9999859,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0164323268701983,"score_gpt":0.2841574089856457,"score_spread":0.2677250821154474,"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."}}