Homepage
Huiyi Chen 陈慧仪
👋 I am a first-year CS Ph.D. student (Aug. 2026–) at 
University of Illinois Chicago (UIC), advised by Prof. Baoxin Li. I received my M.Phil. in Artificial Intelligence from 
Southeast University, advised by Prof. Xu Yang, and my B.S. in Network Engineering from 
Nanchang Hangkong University. Previously, I was a research intern at 
Alibaba Group, working on Video Understanding in Multimodal Large Language Models.
My research interests include:
- Multimodal Large Language Models (MLLMs) & Visual Understanding
- In-context Learning (ICL)
- Robustness and Trustworthiness of MLLMs
- World Models
- Medical Imaging
I am always open to research collaborations and internship opportunities. Feel free to reach out!
🎉 News
- [2026.08] Moved to Chicago 🇺🇸 and started my Ph.D. at UIC!
- [2026.06] New paper: Stop When Reasoning Converges: Semantic-Preserving Early Exit for Reasoning Models (PUMA). 📄 Paper / 📦 Code / 🤗 Models & Datasets
- [2026.05] One paper (MVI-Bench) accepted by ICML 2026! 🎉 📄 Paper / 📦 Code / 🤗 Dataset / Covered by 🔥 智猩猩AI
- [2026.01] Started a research internship at Alibaba Group, working on video understanding in MLLMs.
- [2025.11] New paper released: MVI-Bench — A Comprehensive Benchmark for Evaluating Robustness to Misleading Visual Inputs in LVLMs. 📄 Paper / 📦 Code / 🤗 Dataset
- [2025.07] One paper (KeCO) accepted by ACM MM 2025 as Oral! 🎉 📦 Code
- [2024.02] One paper on in-context sequence configuration for VQA accepted by CVPR 2024! 🎉
🎓 Education
Advisor: Baoxin Li
Advisor: Xu Yang
💼 Internship
Research Intern
Focusing on Video Understanding in Multimodal Large Language Models (MLLMs).
🌟 Activities
Presenter, Seoul, South Korea 🇰🇷
Presented MVI-Bench at the WiML Symposium co-located with ICML 2026.
📝 Publications and Preprints
(* denotes equal contribution.)
First-author Publications
Huiyi Chen, Jiawei Peng, Dehai Min, Changchang Sun, Kaijie Chen, Yan Yan, Xu Yang, Lu Cheng. MVI-Bench: A Comprehensive Benchmark for Evaluating Robustness to Misleading Visual Inputs in LVLMs ICML 2026
📄 Paper / 📦 Code/ 🤗 Dataset /
Huiyi Chen, Jiawei Peng, Kaihua Tang, Xin Geng, Xu Yang. Enhancing Multimodal In-Context Learning for Image Classification through Coreset Optimization ACM MM 2025, Oral
📄 Paper / 📦 CodeLi Li*, Jiawei Peng*, Huiyi Chen*, Chongyang Gao, Xu Yang. How to Configure Good In-Context Sequence for Visual Question Answering CVPR 2024
📄 Paper / 📦 Code
Collaborative Publications
- Dehai Min*, Giovanni Vaccarino*, Huiyi Chen, Yongliang Wu, Gal Yona, Lu Cheng. Stop When Reasoning Converges: Semantic-Preserving Early Exit for Reasoning Models Preprint
📄 Paper / 📦 Code/ 🤗 Models & Datasets
🧑⚖️ Academic Service
- Conference Reviewer: CVPR 2026, NeurIPS 2026, AAAI 2027, WACV 2027
🏅 Awards & Honors
- Southeast University Second-Class Academic Scholarship (Top 10%), 2023-2026
- University-level First-Class Scholarship (Top 1%), 2019-2023
- Outstanding Student Award (Top 1%), 2019-2023
🧩 Hobbies & Interests
Outside research, you will usually find me with a book, an iPad full of half-finished drawings, or a Pixar film I have already seen three times.
📚 Books — I am a devoted reader of Fredrik Backman, and Britt-Marie Was Here is the book I keep coming back to. Britt-Marie, a homemaker nobody ever noticed, wakes up one day and decides to drive alone to Paris. Her story keeps nudging me to be “the kind of person who jumps from the rock into the sea”, and reminds me that it is never too late to start over.
🎬 Films — I am a die-hard
& 
fan. I have watched nearly every classic, plus most of the behind-the-scenes documentaries. Turning Red, Soul, and Inside Out are my top three. One line from Turning Red has stayed with me ever since, when Meimei’s mother tells her: “The farther you go, the prouder I will be.”🎨 Drawing — I draw cute little characters on Procreate whenever I need a break. A few of them live in my Album.