About Me

Hi, my name is Lai Wei (魏来) and I am a second-year PhD student at SJTU, majoring in computer science. Since AI is a subject that interests me greatly, I have been working on it throughout my academic journey. I am committed to staying abreast of new advances in AI and am always looking for novel approaches to innovate and push the boundaries in this field. My recent research focuses on the continual reinforcement learning of (multimodal) large language models.

Educations

Shanghai Jiao Tong University
PhD of Computer Science (2024 - present)
Bachelor of Physics (2020 - 2024)

2025 Young Scientists Sponsorship Program by CAST (中国科协青年科技人才培育工程博士生专项计划)
2025 Zhongguancun Academy Scholarship (top 1.5%)
2025 Shanghai Jiao Tong University Merit Student
2024 Shanghai Jiao Tong University Outstanding Graduates
2022 Shanghai Jiao Tong University Huawei Scholarship
2022 Shanghai Jiao Tong University Undergraduate B Scholarship
2021 Shanghai Jiao Tong University, School of Physics and Astronomy, First Prize Scholarship
2021 Shanghai Jiao Tong University Undergraduate B Scholarship
2021 China Undergraduate Mathematical Contest in Model, Shanghai Division, First Prize

*Equal contribution

First SFT, Second RL, Third UPT: Continual Improving Multi-Modal LLM Reasoning via Unsupervised Post-Training
Lai Wei, Yuting Li, Chen Wang, Yue Wang, Linghe Kong, Weiran Huang, Lichao Sun
[code] [机器之心]
NeurIPS 2025
Diabetica: Adapting Large Language Model to Enhance Multiple Medical Tasks in Diabetes Care and Management
Lai Wei *, Zhen Ying *, Muyang He *, Yutong Chen, Qian Yang, Yanzhe Hong, Jiaping Lu, Kaipeng Zheng, Shaoting Zhang, Xiaoying Li, Weiran Huang, Ying Chen
[code] [model] [ScienceAI]
SCI-FM @ ICLR 2025
Diff-eRank: A Novel Rank-Based Metric for Evaluating Large Language Models
Lai Wei *, Zhiquan Tan *, Chenghai Li, Jindong Wang, Weiran Huang
[code] [量子位]
NeurIPS 2024

*Equal contribution

Advancing Multimodal Reasoning via Reinforcement Learning with Cold Start
Lai Wei, Yuting Li, Kaipeng Zheng, Chen Wang, Yue Wang, Linghe Kong, Lichao Sun, Weiran Huang
[code]
InstructionGPT-4: A 200-Instruction Paradigm for Fine-Tuning MiniGPT-4
Lai Wei, Zihao Jiang, Weiran Huang, Lichao Sun
[code] [机器之心] [video]