4 papers
SafeNexus: Discovering and Steering Modality-Universal Safety Neurons in MLLMs
Jian Yu, Fei Shen, Cong Wang +6
Although Large Language Models (LLMs) have demonstrated promising safety performance, extending them to Multimodal Large Language Models (MLLMs) exposes a significant gap between e…
Who Transfers Safety? Identifying and Targeting Cross-Lingual Shared Safety Neurons
Xianhui Zhang, Chengyu Xie, Linxia Zhu +6
Multilingual safety remains significantly imbalanced, leaving non-high-resource (NHR) languages vulnerable compared to robust high-resource (HR) ones. Moreover, the neural mechanis…
TraceRouter: Robust Safety for Large Foundation Models via Path-Level Intervention
Chuancheng Shi, Shangze Li, Wenjun Lu +5
Despite their capabilities, large foundation models (LFMs) remain susceptible to adversarial manipulation. Current defenses predominantly rely on the "locality hypothesis", suppres…
Where Culture Fades: Revealing the Cultural Gap in Text-to-Image Generation
Chuancheng Shi, Shangze Li, Shiming Guo +9
Multilingual text-to-image (T2I) models have advanced rapidly in terms of visual realism and semantic alignment, and are now widely utilized. Yet outputs vary across cultural conte…