{"id":"W4403024522","doi":"10.1109/pacrim61180.2024.10690230","title":"Exploiting LLM Embeddings for Content-Based IoT Anomaly Detection","year":2024,"lang":"en","type":"article","venue":"","topic":"Network Security and Intrusion Detection","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Victoria","funders":"","keywords":"Computer science; Anomaly detection; Internet of Things; Content (measure theory); Anomaly (physics); Artificial intelligence; Computer security; Mathematics","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":[0.000297542,0.0001013898,0.00009230428,0.000140537,0.000190869,0.0003758566,0.0002177643,0.0000654554,0.00003711944],"category_scores_gemma":[0.00004209828,0.00009090945,0.000111441,0.0003717988,0.00001541763,0.0002761238,0.00005039979,0.0001073369,0.0000524309],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006394364,"about_ca_system_score_gemma":0.00002974875,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006454959,"about_ca_topic_score_gemma":0.00007217282,"domain_scores_codex":[0.9990774,0.00002357768,0.0001859436,0.0003517194,0.000139958,0.0002213464],"domain_scores_gemma":[0.9994878,0.0001654364,0.00003299268,0.0001819757,0.00007545264,0.00005637302],"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":[0.00003846104,0.00006117021,0.00002701091,0.0001303615,0.00002932822,0.00001033149,0.0004221914,0.0004945584,0.09784433,0.06025095,0.001329771,0.8393615],"study_design_scores_gemma":[0.0001569875,0.0002169928,0.00002963022,0.00004491655,0.000004714325,0.000009397827,0.00003889977,0.7899225,0.1932058,0.00112894,0.01512452,0.0001166376],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1581025,0.0001218631,0.8383824,0.0005667061,0.001284384,0.0001822917,8.803395e-7,0.0007562502,0.0006026765],"genre_scores_gemma":[0.9778434,0.000002432572,0.02080677,0.0005180528,0.0002722129,0.00006490512,9.569426e-7,0.00001082609,0.0004804068],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8392449,"threshold_uncertainty_score":0.370718,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04034285777715715,"score_gpt":0.2580560271588107,"score_spread":0.2177131693816535,"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."}}