Large Language Models of Code Fail at Completing Code with Potential Bugs

Tuan Dinh, Jinman Zhao, Samson Tan, Renato Negrinho, Leonard Lausen, Sheng Zha, George Karypis