Zhaoyu Li (李照宇)


  • Jan 1, 2023 -- Transfer to the University of Toronto! New life and new challenges!
  • Sep 14, 2022 -- Our paper "NSNet: A General Neural Probabilistic Framework for Satisfiability Problems" is accepted to NeurIPS 2022! Both paper and code are released.
  • Sep 1, 2021 -- Start my Ph.D. career at McGill University and Mila!
  • Jul 24, 2021 -- Invited talk on "Graph Contrastive Pre-training for Effective Theorem Reasoning", on ICML 2021 Workshop: Self-Supervised Learning for Reasoning and Perception.
  • Jul 13, 2021 -- One paper get accepted by ICML 2021 Workshop: Self-Supervised Learning for Reasoning and Perception as an oral presentation!
  • Jun 27, 2021 -- Graduate from Shanghai Jiao Tong University!



Photo credit: Weizhe Chen.

I am a Ph.D. student in Computer Science at the University of Toronto, working with Prof. Xujie Si. Before coming to U of T, I have spent one year and a half at McGill University and Mila - Quebec AI Institute as a Ph.D. student. Prior to that, I obtained my bachelor's degree with honors from ACM Honors class, Shanghai Jiao Tong University (SJTU).

During my undergraduate studies, I have been fortunate enough to work with Prof. Junchi Yan as a research assistant at SJTU Think Lab in 2019, with Prof. Le Song as a research intern at Gatech Machine Learning Group in 2020.

My research interest mainly lies in neuro-symbolic learning and neural logical reasoning. Recently, I focus on using neural networks to generate symbolic rules, working on the automated theorem proving problem in both formal and informal settings.

With theoretical understanding and specific domain knowledge, I am also interested in developing machine learning algorithms for satisfiability problems, e.g., SAT and #SAT (model counting) problems, and applying them as reasoning engines in some downstream applications.

If you find any research interests that we might share, feel free to drop me an email. I am always open to potential collaborations.


  1. G4SATBench: Benchmarking and Advancing SAT Solving with Graph Neural Networks
    Zhaoyu Li, Jinpei Guo, Xujie Si
  2. Learning Reliable Logical Rules with SATNet
    NeurIPS 2023
    Zhaoyu Li, Jinpei Guo, Yuhe Jiang, Xujie Si
  3. Neuro-symbolic Learning Yielding Logical Constraints
    NeurIPS 2023
    Zenan Li, Yunpeng Huang, Zhaoyu Li, Yuan Yao, Jingwei Xu, Taolue Chen, Xiaoxing Ma, Jian Lu
  4. NSNet: A General Neural Probabilistic Framework for Satisfiability Problems [code]
    NeurIPS 2022
    Zhaoyu Li, Xujie Si
  5. Graph Contrastive Pre-training for Effective Theorem Reasoning
    ICML Workshop on Self-Supervised Learning for Reasoning and Perception 2021 (Contributed talk)
    Zhaoyu Li, Binghong Chen, Xujie Si

Selected Awards

  • GREAT Awards, McGill University, 2022
  • Max Stern Recruitment Fellowship, McGill University, 2021
  • Graduate Excellence Fellowship, McGill University, 2021 - 2022
  • Chinese National Scholarship, (Top 1% in SJTU), 2020
  • Excellent School-level Scholarship, (Top 3% in SJTU), 2017 - 2019
  • Zhiyuan Honorary Scholarship, (Top 5% in SJTU), 2017 - 2020

Teaching Experience

  • Teaching Assistant of CSC2547: Automated Reasoning with Machine Learning, 2023, U of T
  • Guest Lecturer of MS326: Deep Learning and Its Applications, 2022, SJTU
  • Teaching Assistant of MS110: Computer System (2), 2020, SJTU
  • Teaching Assistant of MS108: Computer System (1), 2019, SJTU


  • zhaoyu [at] cs [dot] toronto [dot] edu