{"id":"W7125643653","doi":"10.1109/bigdatase66491.2025.00018","title":"Woodpecker: A Locally Deployed Large Language Model for Protecting Sensitive Information via RAG and Semantic Recognition","year":2025,"lang":"","type":"article","venue":"","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":false,"ca_institutions":"123 Certification (Canada)","funders":"","keywords":"Language model; Semantics (computer science); Information model; Natural language; Feature (linguistics); Key (lock)","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":[0.001828426,0.0003632821,0.0003962447,0.0004082666,0.0009358719,0.0005359338,0.0003014515,0.000257467,0.000008163141],"category_scores_gemma":[0.00145696,0.0003823089,0.0001153246,0.000683344,0.00006877926,0.00260483,0.0006408115,0.0006699233,0.00002034272],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001395236,"about_ca_system_score_gemma":0.0002678403,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001581127,"about_ca_topic_score_gemma":0.0001405562,"domain_scores_codex":[0.9974934,0.0002842485,0.0007068123,0.0005776359,0.0003014376,0.0006364134],"domain_scores_gemma":[0.998147,0.0004711401,0.000379476,0.000381771,0.0005155765,0.0001050051],"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":[0.0004243645,0.0001227847,0.0001345073,0.001312627,0.0002087283,0.00001377628,0.03981939,0.1216954,0.0008065939,0.006094839,0.0001847272,0.8291823],"study_design_scores_gemma":[0.002337107,0.0001123835,0.00008338344,0.0004382809,0.00009588491,0.00001999108,0.002243388,0.9889657,0.001886102,0.00337693,0.00003309807,0.0004077304],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02132978,0.00004855388,0.9736674,0.001026968,0.0003796708,0.002332926,0.0000211598,0.0003190991,0.0008743982],"genre_scores_gemma":[0.8000213,0.000008620326,0.1983106,0.001061406,0.00006578828,0.00009489325,0.00003283868,0.00001769324,0.0003868203],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8672704,"threshold_uncertainty_score":0.9998629,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01084179017092142,"score_gpt":0.263793288223626,"score_spread":0.2529514980527046,"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."}}