ICSLAB邀请到美国康涅狄格大学的洪源教授作学术报告：An LLM-Assisted Easy-to-Trigger Backdoor Attack on Code-Generation Large Language Models (LLMs)