@misc{indiciae092bf1f88acd,
  title = {Jailbreaking LLMs via Semantically Relevant Nested Scenarios with Targeted Toxic Knowledge},
  author = {Ning Xu and Bo Gao and Hui Dou},
  year = {2025},
  url = {https://arxiv.org/abs/2510.01223},
  note = {Source identifier: 2510.01223}
}