PhD Student at the intersection of ML and Physics
Exceeding double-hybrid accuracy at meta-GGA cost by transferring chemical accuracy from wavefunction methods to DFT. A performant JAX pipeline for machine-learned exchange-correlation functionals and Hamiltonian prediction for SCF acceleration, bundling my PhD research into one actively maintained package.
MESS: Modern Electronic Structure Simulations
NequIP is a code for building E(3)-equivariant interatomic potentials
Differentiable, Hardware Accelerated, Molecular Dynamics
Guided Machine Learning Project I created and supervised for undergraduate students of phyics.
A HEP Program for Flavor Observables