ABOUT

I am a research scientist at Meta FAIR. I work at the intersection of machine learning, computational chemistry, and materials science. My research aims to leverage the multi-scale and multi-modal nature of physical systems, as well as their symmetry properties, to develop powerful and scalable learning models and algorithms. My research combines generative modeling and learned simulation techniques to accelerate the design of new materials and molecules with desired properties. I completed my PhD at MIT CSAIL, advised by Tommi Jaakkola.

NEWS

• 25/02 We bridge the gap between test-set accuracy and physical property prediction performance of MLIPs in our new preprint. We propose eSEN, a model that achieves SOTA performance on (compliant) Matbench-Discovery and the MDR Phonon benchmark.
• 25/01 MatterGen is published in Nature. Code is available under an MIT license.
• 24/10 The FAIR Chemistry team released the OMat24 dataset, one of the largest open dataset for materials quantum mechanical calculations.