About

Hi! I am Yunxiang Wei (魏云翔 in Chinese), an undergraduate student at Zhejiang University.

My primary research interest lies in CodeLLMs and Agents. I am particularly interested in deploying LLMs in real-world tasks such as coding and scientific research.

In coding, my work focuses on the reliability of AI-generated code. In scientific discovery, I am interested in equipping AI systems with interdisciplinary knowledge to support idea generation, evaluation, and verification.

My long-term goal is to build automated systems with continual knowledge learning capability as well as reliability.

Publications

Are AI Coders Snitches?

Are AI Coders Snitches? An Empirical Study of Training Data Detection on Code Large Language Models

Tianlin Li, Yunxiang Wei, Zhiming Li, Aishan Liu, Qing Guo, Xianglong Liu, Dongning Sun, Yang Liu

arXiv, 2025.

arXiv / Code

Predicting Program Correctness With Ensemble Semantic Entropy

Predicting Program Correctness With Ensemble Semantic Entropy

Yunxiang Wei, Tianlin Li, Zheng Yuwei, Yanni Dong, Aishan Liu, Xianglong Liu

arXiv, 2026.

arXiv / Code

InnoEval

InnoEval: On Research Idea Evaluation as a Knowledge-Grounded, Multi-Perspective Reasoning Problem

Shuofei Qiao, Yunxiang Wei, Xuehai Wang, Bin Wu, Boyang Xue, Ningyu Zhang, Hossein A. Rahmani, Yanshan Wang, Qiang Zhang, Keyan Ding, Jeff Z. Pan, Huajun Chen, Emine Yilmaz

ICML regular, 2026.

arXiv / Code / Demo

SciAtlas

SciAtlas: A Large-Scale Knowledge Graph for Automated Scientific Research

Shuofei Qiao*, Yunxiang Wei*, Jiazheng Fan, Bin Wu, Busheng Zhang, Mengru Wang, Yuqi Zhu, Ningyu Zhang, Keyan Ding, Qiang Zhang, Huajun Chen

arXiv, 2026.

arXiv / SciAtlas

Education

Zhejiang University

Aug 2023 - Jun 2027 (expected)

B.S. in Computer Science and Technology