 What Was Your Prompt? A Remote Keylogging Attack on AI AssistantsUSENIX Security Symposium (USENIX Security), 2024 |
 Mercury: An Automated Remote Side-channel Attack to Nvidia Deep Learning
AcceleratorInternational Conference on Field-Programmable Technology (ICFPT), 2023 |
 Careful What You Wish For: on the Extraction of Adversarially Trained
ModelsConference on Privacy, Security and Trust (PST), 2022 |
 I Know What You Trained Last Summer: A Survey on Stealing Machine
Learning Models and DefencesACM Computing Surveys (ACM CSUR), 2022 |
 On the Security Risks of AutoMLUSENIX Security Symposium (USENIX Security), 2021 |
 A Protection against the Extraction of Neural Network ModelsInternational Conference on Information Systems Security and Privacy (ICISSP), 2020 |
 Cryptanalytic Extraction of Neural Network ModelsAnnual International Cryptology Conference (CRYPTO), 2020 |
 Quantifying (Hyper) Parameter Leakage in Machine LearningIEEE International Conference on Multimedia Big Data (ICMBD), 2019 |
