@misc{indiciae2df422138b75,
  title = {Security Tensors as a Cross-Modal Bridge: Extending Text-Aligned Safety to Vision in LVLM},
  author = {Shen Li and Liuyi Yao and Wujia Niu and Lan Zhang and Yaliang Li},
  year = {2025},
  url = {https://arxiv.org/abs/2507.20994},
  note = {Source identifier: 2507.20994}
}