About

I am Yongqiang Tian, a Lecturer (Assistant Professor equivalent) in software engineering at Monash University. My research focuses on software testing and debugging for reliable and trustworthy AI-enabled software systems, with particular interests in validation, error detection, and automated techniques for reliable and reproducible computing.

I have academic, research, and teaching experience across Hong Kong, Canada, and Australia. Across these settings, I have worked on deep learning systems, compilers, code intelligence, and other technically demanding software workflows where reliability matters.

A compact overview of how my research connects systematic methods, modern software systems, and distinctive contributions and real-world impact.

Two Questions I Care About

Within this broader roadmap, there are two research questions that I keep coming back to:

• What important properties of software should be tested, but are still not well studied? Representative examples:
  • LPO: Discovering Missed Peephole Optimizations with Large Language Models (ASPLOS 2026)
  • To What Extent Do DNN-based Image Classification Models Make Unreliable Inferences? (EMSE 2021)
  • An Empirical Study of Bugs in Data Visualization Libraries (FSE 2025)
  • Unmasking the Type Inference Capabilities of LLMs for Java Code Snippets (TOSEM 2026)
• How can software be tested and debugged effectively and efficiently with respect to those properties? Representative examples:
  • Compilation Consistency Modulo Debug Information (ASPLOS 2023)
  • Optimization-Aware Test Generation for Deep Learning Compilers (ICSE 2026)
  • A Tale of Two DL Cities: When Library Tests Meet Compiler (ICSE 2025)
  • LPR: Large Language Models-Aided Program Reduction (ISSTA 2024)

More

• Publications: Full publication list, selected papers, venue notes, and citation links.
• Openings: Information for prospective Ph.D. students, interns, research assistants, and undergraduate researchers.
• Teaching: Course leadership, student feedback, and teaching experience across Monash, HKUST, and Waterloo.
• Service: ACM SIGSOFT leadership, program committees, journal reviewing, and artifact evaluation.
• Funding & Awards: Grants, fellowship schemes, industry support, and recognition.
• Profiles and Resources: Google Scholar, DBLP, ORCID, LaTeX writing guide, awesome-research collection, and DBLP BibTeX copier tool.