Mitigating Backdoor Threats to Large Language Models: Advancement and Challenges | IEEE Conference Publication | IEEE Xplore