About Me

I am a second-year Ph.D. student at the University of Maryland, College Park, advised by Prof. Heng Huang. Previously, I worked as a research assistant at the Natural Language Processing Laboratory of Northeastern University (China), supervised by Prof. Tong Xiao. I received my B.E. degree from Northeastern University of Computer Science and Engineering in 2021.

My research is driven by a core vision: to build AI that is both fundamentally intelligent and universally accessible. I focus on creating novel paradigms that unlock the full potential of (large) language models, covering three key areas:

• Efficient Backbone/Foundation Model: Multi-scale Modeling (UMST, ICML2022), Subspace Interaction (EIT, ACL2024), Subspace Modeling (PartialFormer, ACL2024), ODE (PCFormer, NeurIPS2024), Encode-Decoder Trade-offs (LaMaTE, ACL2025)
• Efficient Training and Inference Strategies: Multi-Draft Speculative Decoding (ICLR2025), Asymmetric Training for MMT (ACL2025), Self-evolving Training (Arxiv)
• Next-Generation Reasoning Paradigms: Developing scalable RL frameworks for efficient parallel reasoning, such as Formal-Informal Language Co-Evolving Training and inference (Arxiv), Parallel Thinking (Arxiv)