{"id":"W4405717542","doi":"10.1109/tdsc.2024.3521451","title":"Feature Reconstruction Attacks and Countermeasures of DNN Training in Vertical Federated Learning","year":2024,"lang":"en","type":"article","venue":"IEEE Transactions on Dependable and Secure Computing","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Training (meteorology); Feature (linguistics); Computer security; Feature extraction; Artificial intelligence; Multimedia","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":[0.00474357,0.0009153216,0.000899869,0.001023244,0.0009468857,0.001242852,0.001786808,0.002214166,0.001439043],"category_scores_gemma":[0.02020433,0.0004808128,0.0006310729,0.0006927734,0.001636396,0.00360431,0.003175219,0.00281887,0.0004356239],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001665271,"about_ca_system_score_gemma":0.001728256,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002064516,"about_ca_topic_score_gemma":0.001729758,"domain_scores_codex":[0.9954482,0.001390995,0.0003321842,0.0008468831,0.001356075,0.0006257198],"domain_scores_gemma":[0.9898974,0.004355996,0.0008163966,0.003697295,0.0009023072,0.0003307389],"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":[0.003025727,0.0006279036,0.01065737,0.0001818143,0.0002540514,0.0006592452,0.0003630057,0.4321196,0.0196012,0.05819125,0.009389519,0.4649293],"study_design_scores_gemma":[0.00003297977,0.0001220128,0.0007462921,0.00002659597,0.00001726158,0.0001421381,0.00004624617,0.9706654,0.009585309,0.01770908,0.0008927693,0.00001394491],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3028529,0.0008290316,0.6777465,0.003462211,0.0003432689,0.0002042534,0.0002702354,0.005666632,0.008624846],"genre_scores_gemma":[0.9534712,0.00008015703,0.04460492,0.0003445565,0.00003415258,0.00004974301,0.0001569575,0.00004713708,0.001211146],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.00474357,"threshold_uncertainty_score":0.0250867,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01503552347500677,"score_gpt":0.2595993449859744,"score_spread":0.2445638215109676,"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."}}