Wei-Jie Xu

Currently I am a master student of School of Artificial Intelligence in Nanjing University and a member of LAMDA group, led by Prof. Zhi-Hua Zhou.

In June 2024, I received my B.Eng. degree from School of Computer Science and Technology, Soochow University. In the same year, I was admitted to study for a master's degree in School of Artificial Intelligence, Nanjing University under the supervision of Prof. Kai Ming Ting.

Email  /  Github  /  Google Scholar  /  DBLP  /  LinkedIn

profile photo

News

  • [Aug.2025] LLM4MS has been accepted by Communication Chemistry (SCI I) !
  • [Aug.2025] PIP has been accepted as Findings of EMNLP2025 !
  • [Jan.2025] Released PIP (Perturbation-based Iterative Pruning for LLMs).
  • [Nov.2024] Released DBSC.
  • [Oct.2024] Released D-Spec.
  • [Sep.2024] Enrolled as a master student in Nanjing University.

Research

Research is driven by curiosity and enjoyment, though practical implications may emerge. I have varying degrees of interest in all aspects of artificial intelligence.

PIP Paper Image
PIP: Perturbation-based Iterative Pruning for Large Language Models
Yi Cao, Wei-Jie Xu, Yucheng Shen, Weijie Shi, Chi-Min Chan, Jiajie Xu
EMNLP2025 Findings

PIP (Perturbation-based Iterative Pruning), a novel double-view structured pruning method to optimize LLMs, which combines information from two different views: the unperturbed view and the perturbed view, establishing it as a leading technique for optimizing LLMs in environments with constrained resources.

Award

  • New Point Software Scholarship,[link],2021-2022, School of Computer Science and Technology, Soochow University.

Experience

alibaba logo
Algorithm Intern
Quark, Alibaba Group, Jul.2025 - Oct.2025
Deep Search, Reinforcemen Learning, VeRL
  • Quark, Intelligent Information, Alibaba Group
  • Built and Improved the post-training framework for agentic rl
  • Assisted in building a data synthesis framework
    		•