My Research

Graph Neural Networks and Reinforcement Learning for Quantum Simulation

** Update ** Our submission involving this work was accepted for The 4th Workshop on Mathematical Reasoning and AI at NeurIPS 2024! 🥳
Preprint to appear.

I am working with Richie Yeung and Matthew Sutcliffe on leveraging Graph Neural Networks (GNNs) and Reinforcement Learning (RL) to enhance quantum simulation. This interdisciplinary approach aims to improve the efficiency of simulating quantum circuits on classical computers, which is crucial for testing quantum software without relying heavily on quantum computing resources that are currently scarce and expensive.

View on GitHub (Not yet publicly available)

Hybrid Quantum Machine Learning with String Diagrams

I am developing a formal framework for hybrid quantum-classical machine learning algorithms, leveraging category theory—a branch of pure mathematics. This facilitates the design of programming languages tailored for hybrid algorithms and reveals key details about the interaction between classical machine learning and quantum computing, helping us reason about these algorithms. An elegant feature of category theory is its capacity to represent quantum machine learning algorithms through diagrams that are mathematically rigorous, yet intuitively comprehensible.

View my first paper on this here!

Previous Projects

Previously, I have also worked on an interactive theorem prover for graphical languages as well as grammar-aware language models.