I am a fifth-year Ph.D. candidate in the School of Data Science at the University of Virginia, advised by Prof. Sheng Li. Before joining UVA, I completed my master’s studies at the University of Georgia and at the National University of Singapore. I received my bachelor’s degree from Northeastern University in China, and prior to pursuing graduate studies in the United States, I worked as a C/C++ engineer in Chengdu.

My research primarily focuses on

• Large Language Model Reasoning
• Graph Representation Learning
• Trustworthy Machine Learning
• AI for good

I am always open to collaborating on interesting and impactful projects. Feel free to reach out to me at weili.shi [at] virginia.edu.

News

Recent Publications

• Finding the Cracks: Improving LLMs Reasoning with Paraphrastic Probing and Consistency Verification (under review)
  Weili Shi, Dongliang Guo, Lehan Yang, Tianlong Wang, Hanzhang Yuan, Sheng Li.

• LLMs Are Too Smart to Be Average: Controlling LLM Proficiency via Guided Decoding (under review)
  Dongliang Guo, Weili Shi, Sheng Li.

• Revisiting Dynamic Graphs from the Perspective of Time Series (under review)
  Weili Shi, Dongliang Guo, Sheng Li.

• Leveraging Large Language Models for Dynamic Text-Attributed Graphs Learning (under review)
  Weili Shi, Haoming Li, Sheng Li.

• Dual-windowed Vision Transformer with Angular Self-Attention
  Transactions on Machine Learning Research (TMLR), 2024.
  Weili Shi, Sheng Li.

Awards

• Notable reviewer from The International Conference on Learning Representations (ICLR) 2025.
• Review Certificate from IEEE Transactions on Knowledge and Data Engineering (TKDE) 2024.

Hobbies

I enjoy swimming, tennis, table tennis, badminton, hiking, and other outdoor activities.