Chenguang S. Wang

I am currently a Presidential Postdoctoral Fellow at Virginia Tech, working primarily under the supervision of Prof. Dawei Zhou, with co-advisors Prof. Lu Zhang and Prof. Le Wang. My research develops harnesses and agentic AI systems for understanding, improving, and deploying Large Language Models (LLMs), with an emphasis on interpretable reasoning, robust decision support, and high-stakes real-world applications such as assessment, infrastructure design, and disaster management.

I received my Ph.D. Degree from Stony Brook University, where I was advised by Prof. Susu Xu. Prior to that, I earned a Master of Science in Computer Science from Stevens Institute of Technology, where I worked with Prof. Xiaojiang Du, and a Bachelor of Science in Computer Science from Xi’an Jiaotong University. From 2024 to 2025, I was also a Graduate Visiting Scholar at Johns Hopkins University.

My research interests broadly lie in Machine Learning (ML), Natural Language Processing (NLP), and Large Language Models (LLMs).
More specifically, my recent research focuses on:

• LLM evaluation and control harnesses for high-stakes applications (Epi2Diff, Item Discrimination, RuleR, Mosaic-IT)

• Agentic AI systems for decision support, disaster response, and design (FLARE, CrisisNLP, StreetDesigner)

• Multi-modal reasoning and trustworthy MLLM evaluation (TRIG, CaughtCheating)

Across these directions, I aim to build AI systems that are not only capable, but also interpretable, controllable, and useful in settings where reliability matters. I welcome collaborations on trustworthy LLM systems, agentic workflows, and high-impact AI applications across academic and industry settings.