DRIP: Defending Prompt Injection via Token-wise Representation Editing and Residual Instruction Fusion

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Liu, Ruofan, Lin, Yun, Huang, Zhiyong, Dong, Jin Song
Format:	Preprint
Published:	2025
Subjects:	Cryptography and Security Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2511.00447