Shaonan Wu (武少楠)

Google Scholar | LinkedIn

Please contact me with this email: shaonanwu[AT]stu.xjtu.edu.cn

Shaonan Wu is a second-year Ph.D. student in the School of Artificial Intelligence at Xi'an Jiaotong University (XJTU) and a research intern at the Vision and Multimodal Lab, JD Explore Academy. Previously, he was a joint Ph.D. student with Microsoft Research Asia and XJTU, co-supervised by Dr. Ping Wei and Dr. Nan Duan. He received his B.S. degree in Artificial Intelligence from XJTU in 2024. His research interests include machine reasoning and foundation models.

Research Experience

• [2023/11 ~ 2024/9] Research Intern, Microsoft Research Asia (Mentor: Shuai Lu & Nan Duan)
• [2024/9 ~ 2025/11] Research Intern & Joint PhD student, Microsoft Research Asia (Mentor: Shuai Lu & Yeyun Gong)
• [2025/11 ~ current] Research Intern, JD Explore Academy

Publication

• Alchemy: Amplifying Theorem-Proving Capability through Symbolic Mutation
  Shaonan Wu*, Shuai Lu, Yeyun Gong, Nan Duan, Ping Wei.
  ICLR 2025.
  TL;DR: We propose a CPU-only method to synthesize new theorems based on existing theorems, which shows promise in enhancing the theorem-proving ability of neural provers.

Selected Honors

• [2024/6] Outstanding graduates of XJTU.