SOURCE-LINKED INTELLIGENCE

Do VLMs Share Safety Neurons Across Modalities?

arXiv · AI, language, vision and robotics · article · Aug 31, 2026 · UTC

Vision-language models (VLMs) can comply with harmful requests delivered through images, even when their LLM backbones would refuse the same content in text. While prior work characterizes these jailbreaks empirically or at the representation level, how visual inputs perturb safety pathways at the neuron level remains uncharted. We close this gap with a causal, neuron-level analysis of safety mechanisms in 10 VLMs. We propose a two-stage detection pipeline with iterative ablation that accounts for self-repair, and introduce two modality-isolated benchmarks, ViSafe-Detect and ViSafe-Eval, which

Read original source ↗ Open in workspace

recordType

paper

region

Global