NeuroArmor: Safe-Variant-Guided Representation Consistency for Selective Re-Anchoring in Jailbreak Defense

Abstract

Large language models remain vulnerable to jailbreak attacks that hide harmful intent behind seemingly ordinary requests such as role-play, translation, encoding, adversarial suffixes, and multi-turn buildup. Existing defenses still struggle to handle these attacks without over-blocking benign but sensitive requests, partly because they often apply the same action to every prompt and therefore fail to balance safety and helpfulness. We propose NeuroArmor, a white-box runtime defense that uses pr

Categories

Framework mappings

Suggested from the entry's categories.

@misc{lin2026neuroarmor,
  title = {{NeuroArmor: Safe-Variant-Guided Representation Consistency for Selective Re-Anchoring in Jailbreak Defense}},
  author = {Zhongyang Lin and Ziran Zhao and Feifei Zhai and Pengyuan Liu},
  year = {2026},
  month = jun,
  eprint = {2606.03486},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2606.03486}
}