Exploring Clean Label Backdoor Attacks and Defense in Language Models | IEEE Journals & Magazine | IEEE Xplore