News

• LogicXGNN: Grounded Logical Rules for Explaining Graph Neural Networks is accepted to ICLR 2026 (Top 2.5%) [Paper].
• Learning Minimal Neural Specifications is accepted to NeuS 2025 (Oral presentation) [Paper].
• Towards Reliable Neural Specifications is accepted to ICML 2023 (Oral presentation) [Paper].
• Identifying different student clusters in functional programming assignments: From quick learners to struggling students is accepted to SIGCSE TS 2023 [Paper].

About me

I am a final-year PhD student currently based at the University of Toronto, where I have been visiting since 2023 after my supervisor, Xujie Si, joined UofT. I am formally affiliated with McGill University and the Montreal Institute for Learning Algorithms (MILA).

My research interests include explainable AI, neuro-symbolic methods, and neural network robustness and verification.

Research interests

• Explainable AI; mechanistic interpretability and circuits;
• Neuro-symbolic methods;
• Neural network robustness and verification;

Education

🎓 PhD in Computer Science (2023-) University of Toronto (visiting)

🎓 PhD in Computer Science (2021-) McGill University / MILA

🎓 MSc in Computer Science (2019-2021) Georgia Institute of Technology

🎓 BSc in Math & Stats (2011-2015) University of Toronto