Jinyao Guo
I am a Ph.D. student in Computer Science at Purdue University, advised by Prof. Xiangyu Zhang. My research focuses on the combination of artificial intelligence and software engineering, especially in the areas of coding models and LLM-assisted program analysis.
AI4SE. I combine program analysis with LLM reasoning about source code and natural-language specifications to detect bugs. My projects include RepoAudit, BugScope, and RFCAudit. RepoAudit audits repositories with millions of lines of code and has identified more than 500 bugs across open-source projects.
SE4AI. I work on agent context management for long-horizon coding tasks and how to generate training data that improves coding models. In ProSec, we used synthetic programming tasks paired with secure solutions to improve the security of generated code while preserving functional correctness. I also in the red team for the Amazon Nova AI Challenge, where I audit agent-generated code for vulnerabilites.