1
DarkLLM: Learning Language-Driven Adversarial Attacks with Large Language Models
提出DarkLLM框架，利用大语言模型自动生成语言驱动的对抗样本，攻破NLP模型防线
arXiv:2605.18868v1 Announce Type: cross Abstract: While vision and multimodal foundation models underpin critical tasks from perception to complex rea…