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

ArXiv CS.AI2026-06-03NEWSen作者: Zhongyang Lin, Ziran Zhao, Feifei Zhai, Pengyuan Liu
查看原文 →
关系图谱

NeuroArmor: Safe-Variant-Guided Representation Consistency for Selective Re-Anchoring in Jailbreak Defense · 相关事件

相关事件

NeuroArmor: Safe-Variant-Guided Representation Consistency for Selective Re-Anchoring in Jailbreak Defense
2026-06-03PRODUCT_LAUNCH影响: MEDIUM