Qingqing Ye

Unlearning Isn't Deletion: Investigating Reversibility of Machine Unlearning in LLMs

Decision Potential Surface: A Theoretical and Practical Approximation of LLM's Decision Boundary

Machine Unlearning in Low-dimensional Feature Subspace

Virus Infection Attack on LLMs: Your Poisoning Can Spread "VIA" Synthetic Data

Toward Efficient Inference Attacks: Shadow Model Sharing via Mixture-of-Experts

Does Low Rank Adaptation Lead to Lower Robustness against Training-Time Attacks?

New Paradigm of Adversarial Training: Breaking Inherent Trade-Off between Accuracy and Robustness via Dummy Classes

Alignment-Aware Model Extraction Attacks on Large Language Models

Analyzing and Optimizing Perturbation of DP-SGD Geometrically