Yiwen Guo

Deciphering the Chaos: Enhancing Jailbreak Attacks via Adversarial Prompt Translation

LLM as a Complementary Optimizer to Gradient Descent: A Case Study in Prompt Tuning

Towards Reliable Backdoor Attacks on Vision Transformers

DualAug: Exploiting Additional Heavy Augmentation with OOD Data Rejection

Improved Generation of Adversarial Examples Against Safety-aligned LLMs

Code Representation Pre-training with Complements from Program Executions

Towards Reliable Backdoor Attacks on Vision Transformers

Characterizing Robust Overfitting in Adversarial Training via Cross-Class Features

The Program Testing Ability of Large Language Models for Code

Ref-Diff: Zero-shot Referring Image Segmentation with Generative Models