Offline Textual Adversarial Attacks against Large Language Models | IEEE Conference Publication | IEEE Xplore