UEval: A Benchmark for Unified Multimodal Generation
ArXiv, January 2026
MSCS Student at UIUC
I work on generative multimodal models, unified multimodal systems, and evaluations for perception and reasoning. My recent research centers on making multimodal models see more faithfully, reason more structurally, and generate with stronger grounding.
I received a lot of help during the Fall 2026 application cycle, and I would love to pass that support on. If you are a junior undergraduate student and would like to chat about applications, summer research, life, career planning, or AI-related research ideas, feel free to reach out.
Selected recent updates.
A cleaner academic structure, now using your own background and research profile.
Hi! I'm Bo Li, an incoming MSCS student at UIUC. Previously, I was a research intern at Princeton University advised by Prof. Zhuang Liu. During this internship, I had the privilege of working closely with amazing collaborators, including Yida Yin, Wenhao Chai, and Dr. Xingyu Fu.
Before that, I worked as a research assistant at UNC-Chapel Hill advised by Huaxiu Yao, and at Fudan University advised by Jingjing Chen.
My research interests include, but are not limited to, AI alignment, unified models for multimodal understanding and generation, and benchmarks and evaluation. My pronouns are She/Her/Hers.
Current selected publications.
Selected honors.
Reviewing and teaching experience.
Small details outside research.
Enjoy playing the piano, and used to play the guitar, though it has been a long time since I last picked it up.
Genshin Impact and Zenless Zone Zero.
Hope to have the chance to travel around the world.