Too Easily Fooled? Prompt Injection Breaks LLMs on Frustratingly Simple Multiple-Choice Questions

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Guo, Xuyang, Huang, Zekai, Song, Zhao, Zhang, Jiahao
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/2508.13214