Zhengyue Zhao

ARMOR: Aligning Secure and Safe Large Language Models via Meticulous Reasoning

PRISM: Robust VLM Alignment with Principled Reasoning for Integrated Safety in Multimodality

VLA-Risk: Benchmarking Vision-Language-Action Models with Physical Robustness

Can Watermarks be Used to Detect LLM IP Infringement For Free?

Adversarial Contrastive Decoding: Aligning Large Language Models via Exploiting Their Safety and Harm