12 papers
Dual-Adversarial Safety Alignment: Cultivating Intrinsic Threat Comprehension in LRMs
Hongli Shen, Shaopeng Fu, Qinbo Zhang +2
Large reasoning models (LRMs) achieve remarkable success on complex tasks but remain vulnerable to harmful prompts that induce unsafe outputs. Recent methods align LRMs using direc…
Benign Overfitting in Adversarial Training for Vision Transformers
Jiaming Zhang, Meng Ding, Shaopeng Fu +2
Despite the remarkable success of Vision Transformers (ViTs) across a wide range of vision tasks, recent studies have revealed that they remain vulnerable to adversarial examples,…
Understanding and Improving Continuous Adversarial Training for LLMs via In-context Learning Theory
Shaopeng Fu, Di Wang
Adversarial training (AT) is an effective defense for large language models (LLMs) against jailbreak attacks, but performing AT on LLMs is costly. To improve the efficiency of AT f…
RefineRL: Advancing Competitive Programming with Self-Refinement Reinforcement Learning
Shaopeng Fu, Xingxing Zhang, Li Dong +2
While large language models (LLMs) have demonstrated strong performance on complex reasoning tasks such as competitive programming (CP), existing methods predominantly focus on sin…
Concept-Based Dictionary Learning for Inference-Time Safety in Vision Language Action Models
Siqi Wen, Shu Yang, Shaopeng Fu +3
Vision Language Action (VLA) models close the perception action loop by translating multimodal instructions into executable behaviors, but this very capability magnifies safety ris…
Accelerating Suffix Jailbreak attacks with Prefix-Shared KV-cache
Xinhai Wang, Shaopeng Fu, Shu Yang +3
Suffix jailbreak attacks serve as a systematic method for red-teaming Large Language Models (LLMs) but suffer from prohibitive computational costs, as a large number of candidate s…
Understanding the Impact of Differentially Private Training on Memorization of Long-Tailed Data
Jiaming Zhang, Huanyi Xie, Meng Ding +3
Recent research shows that modern deep learning models achieve high predictive accuracy partly by memorizing individual training samples. Such memorization raises serious privacy c…
Understanding Private Learning From Feature Perspective
Meng Ding, Mingxi Lei, Shaopeng Fu +3
Differentially private Stochastic Gradient Descent (DP-SGD) has become integral to privacy-preserving machine learning, ensuring robust privacy guarantees in sensitive domains. Des…
Backdooring CLIP through Concept Confusion
Lijie Hu, Junchi Liao, Weimin Lyu +5
Backdoor attacks pose a serious threat to deep learning models by allowing adversaries to implant hidden behaviors that remain dormant on clean inputs but are maliciously triggered…
Short-length Adversarial Training Helps LLMs Defend Long-length Jailbreak Attacks: Theoretical and Empirical Evidence
Shaopeng Fu, Liang Ding, Jingfeng Zhang +1
Jailbreak attacks against large language models (LLMs) aim to induce harmful behaviors in LLMs through carefully crafted adversarial prompts. To mitigate attacks, one way is to per…
Pre-trained Encoder Inference: Revealing Upstream Encoders In Downstream Machine Learning Services
Shaopeng Fu, Xuexue Sun, Ke Qing +2
Pre-trained encoders available online have been widely adopted to build downstream machine learning (ML) services, but various attacks against these encoders also post security and…
Theoretical Analysis of Robust Overfitting for Wide DNNs: An NTK Approach
Shaopeng Fu, Di Wang
Adversarial training (AT) is a canonical method for enhancing the robustness of deep neural networks (DNNs). However, recent studies empirically demonstrated that it suffers from r…