10 papers
TwinGate: Stateful Defense against Decompositional Jailbreaks in Untraceable Traffic via Asymmetric Contrastive Learning
Bowen Sun, Chaozhuo Li, Yaodong Yang +2
Decompositional jailbreaks pose a critical threat to large language models (LLMs) by allowing adversaries to fragment a malicious objective into a sequence of individually benign q…
PRISM: Robust VLM Alignment with Principled Reasoning for Integrated Safety in Multimodality
Nanxi Li, Zhengyue Zhao, G. Edward Suh +2
Safeguarding vision-language models (VLMs) is a critical challenge, as existing methods often suffer from over-defense, which harms utility, or rely on shallow alignment, failing t…
Sample-Specific Noise Injection For Diffusion-Based Adversarial Purification
Yuhao Sun, Jiacheng Zhang, Zesheng Ye +2
Diffusion-based purification (DBP) methods aim to remove adversarial noise from the input sample by first injecting Gaussian noise through a forward diffusion process, and then rec…
ReasoningBomb: A Stealthy Denial-of-Service Attack by Inducing Pathologically Long Reasoning in Large Reasoning Models
Xiaogeng Liu, Xinyan Wang, Yechao Zhang +5
Large reasoning models (LRMs) extend large language models with explicit multi-step reasoning traces, but this capability introduces a new class of prompt-induced inference-time de…
SudoLM: Learning Access Control of Parametric Knowledge with Authorization Alignment
Qin Liu, Fei Wang, Chaowei Xiao +1
Existing preference alignment is a one-size-fits-all alignment mechanism, where the part of the large language model (LLM) parametric knowledge with non-preferred features is unifo…
Benchmarking Vision Language Model Unlearning via Fictitious Facial Identity Dataset
Yingzi Ma, Jiongxiao Wang, Fei Wang +10
Machine unlearning has emerged as an effective strategy for forgetting specific information in the training data. However, with the increasing integration of visual data, privacy c…