• Theoretical physics: Coming up with theories and calculating their predictions is challenging. Could you teach an AI agent to help work out the details? We design symbolic AI algorithms to assist in theoretical physics calculations and build new phenomenological models. We are in the process of significantly expanding this research direction.
• Statistical methods in particle physics and astrophysics: Raw data from experiments or telescopes are complex and high-dimensional, but compressing the data results in information loss. We design fast and high-dimensional statistical inference techniques that have transformed Higgs physics at the Large Hadron Collider, and can transform neutrino physics or neutron star astrophysics. Learn more about our work in particle physics and astrophysics.
• Accelerating discovery: Designing mathematical models that describe Nature requires simulating their predictions and comparing these to data. We develop fast and generalizable machine learning methods to simulate astrophysical and particle physics phenomena. We also develop machine learning tools that could help design future experiments.

Our group engages in cross-disciplinary collaborations and welcomes students from across academic backgrounds. Most projects in the group involve computational tools and AI/ML, so students with strong skills or interest in these areas will find a natural fit here. We are also interested in collaborations in the history of science and scientific outreach. The group welcomes students who have taken career breaks or followed non-traditional career paths and students from across socio-economic backgrounds and genders.