Kangjie Chen

Unmasking Backdoors: An Explainable Defense via Gradient-Attention Anomaly Scoring for Pre-trained Language Models

Picky LLMs and Unreliable RMs: An Empirical Study on Safety Alignment after Instruction Tuning

Analogy-based Multi-Turn Jailbreak against Large Language Models

Transstratal Adversarial Attack: Compromising Multi-Layered Defenses in Text-to-Image Models

TRUST-VLM: Thorough Red-Teaming for Uncovering Safety Threats in Vision-Language Models

Impact-driven Context Filtering For Cross-file Code Completion

REPOFILTER: Adaptive Retrieval Context Trimming for Repository-Level Code Completion