9 papers
A Multimodal Automatic Redteaming Evaluation based on Atomic Jailbreak Strategy Decoupling and Combination
Shiji Zhao, Yuxuan Zhou, Chen Xiong +3
Multimodal Large Language Models (MLLMs) have achieved impressive progress in image-text comprehension and generation, yet they remain susceptible to jailbreak attacks that can tri…
OctoPipe: Reducing Pipeline Bubbles for Heterogeneous Models via Co-Optimizing Partitioning, Placement, and Scheduling
Jihu Guo, Tenghui Ma, Wei Gao +6
Pipeline parallelism is widely used to train large language models (LLMs). However, increasing heterogeneity in model architectures exacerbates pipeline bubbles, thereby reducing t…
The Salami Slicing Threat: Exploiting Cumulative Risks in LLM Systems
Yihao Zhang, Kai Wang, Jiangrong Wu +7
Large Language Models (LLMs) face prominent security risks from jailbreaking, a practice that manipulates models to bypass built-in security constraints and generate unethical or u…
Improving Safety Alignment via Balanced Direct Preference Optimization
Shiji Zhao, Mengyang Wang, Shukun Xiong +7
With the rapid development and widespread application of Large Language Models (LLMs), their potential safety risks have attracted widespread attention. Reinforcement Learning from…
Blackbox Dataset Inference for LLM
Ruikai Zhou, Kang Yang, Xun Chen +3
Today, the training of large language models (LLMs) can involve personally identifiable information and copyrighted material, incurring dataset misuse. To mitigate the problem of d…
Why does weak-OOD help? A Further Step Towards Understanding Jailbreaking VLMs
Yuxuan Zhou, Yuzhao Peng, Yang Bai +7
Large Vision-Language Models (VLMs) are susceptible to jailbreak attacks: researchers have developed a variety of attack strategies that can successfully bypass the safety mechanis…