Xinchen Xu

AI4GC Lab

Xinchen Xu

Multimodal LLMModel Merging

About Me

Hi! I'm Xinchen Xu (许馨宸), a Ph.D. student at Zhejiang University, advised by Shengyu Zhang and Fei Wu.

My research focuses on multimodal large language models and model merging — understanding how models can be composed and aligned to handle complex, cross-modal tasks more efficiently.

Feel free to reach out for collaborations!

Research Interests

  • Multimodal LLM — Building and understanding large language models that perceive and reason across text, vision, and other modalities.
  • Model Merging — Exploring how independently trained models can be merged to combine capabilities without expensive retraining.

Selected Papers

ACL2025

OS Agents: A Survey on MLLM-based Agents for Computer, Phone and Browser Use

PaperProject485