Zhongyang Li
Hi there, I am a Master’s student in Computer Science at Johns Hopkins University, advised by Prof. Ashutosh Dutta. Currently, I am a research intern at the University of Maryland, College Park, working with Prof. Tianyi Zhou. My research interests center on efficient foundation models, with a specific focus on Mixture-of-Experts (MoE) and Test-Time Adaptation. My work aims to improve model’s generalization and efficiency.
Seeking PhD Opportunities: I am actively applying for PhD positions for Fall 2026 in the US. If you are interested in my research, please feel free to contact me!
news
| Jan 22, 2026 | 🔥 Super excited to share that our paper “Routing Manifold Alignment Improves Generalization of MoE LLMs” has been accepted to ICLR 2026! |
|---|---|
| Aug 14, 2025 | 🏆 Honored to receive the Travel Award from COLM 2025! I will also serve as a student volunteer at the conference in Montreal. See you there! |
| Jul 12, 2025 | ✈️ I will attend ICML 2025 in Vancouver to present our poster for R2-T2. Come say hi if you are around! |
| Jul 07, 2025 | 🎉 Our paper C3PO (Critical-Layer, Core-Expert, Collaborative Pathway Optimization) has been accepted to COLM 2025! |
| Jun 01, 2024 | 👨💻Started my research internship at the University of Maryland, College Park, advised by Prof. Tianyi Zhou. I will be working on efficient training and inference for MoE models. |
