16 papers
The Art of (Mis)alignment: How Fine-Tuning Methods Effectively Misalign and Realign LLMs in Post-Training
Rui Zhang, Hongwei Li, Yun Shen +6
The deployment of large language models (LLMs) raises significant ethical and safety concerns. While LLM alignment techniques are adopted to improve model safety and trustworthines…
UTOPIA: Unlearnable Tabular Data via Decoupled Shortcut Embedding
Jiaming He, Fuming Luo, Hongwei Li +5
Unlearnable examples (UE) have emerged as a practical mechanism to prevent unauthorized model training on private vision data, while extending this protection to tabular data is no…
BadTemplate: A Training-Free Backdoor Attack via Chat Template Against Large Language Models
Zihan Wang, Hongwei Li, Rui Zhang +2
Chat template is a common technique used in the training and inference stages of Large Language Models (LLMs). It can transform input and output data into role-based and templated…
TEAR: Temporal-aware Automated Red-teaming for Text-to-Video Models
Jiaming He, Guanyu Hou, Hongwei Li +6
Text-to-Video (T2V) models are capable of synthesizing high-quality, temporally coherent dynamic video content, but the diverse generation also inherently introduces critical safet…
Hidden Tail: Adversarial Image Causing Stealthy Resource Consumption in Vision-Language Models
Rui Zhang, Zihan Wang, Tianli Yang +5
Vision-Language Models (VLMs) are increasingly deployed in real-world applications, but their high inference cost makes them vulnerable to resource consumption attacks. Prior attac…
ConfGuard: A Simple and Effective Backdoor Detection for Large Language Models
Zihan Wang, Rui Zhang, Hongwei Li +4
Backdoor attacks pose a significant threat to Large Language Models (LLMs), where adversaries can embed hidden triggers to manipulate LLM's outputs. Most existing defense methods,…