Scenes as an Attack Surface: Uncovering Scene-Grounded Safety Failures in MLLMs
arXiv:2610.14094v1 Announce Type: new Abstract: Existing jailbreak attacks have exposed serious safety weaknesses in multimodal large language models (MLLMs). Most prior attacks rely on specially constructed visual inputs to carry or support harmful intent. In this paper, we uncover a different vul…
Read the full story at arXiv cs.CV ↗
Timeline · 1 report
- 2026-10-14 04:00 · arXiv cs.CV
Scenes as an Attack Surface: Uncovering Scene-Grounded Safety Failures in MLLMs