Research

FLAME

A library for atomistic modeling environments. FLAME is a computational library for developing neural network interatomic potentials, structure prediction, and atomistic simulations.

MALA

Materials Learning Algorithms. A data-driven framework for predicting materials properties using machine learning.

PyFLAME

An automated workflow for developing neural network interatomic potentials with FLAME. Streamlines the process of training and deploying MLIPs.