David Rolnick
McGill University and Mila, Montreal

Professor and Canada CIFAR AI Chair in the School of Computer Science

David Rolnick is an Assistant Professor and Canada CIFAR AI Chair in the School of Computer Science at McGill University and at Mila Quebec AI Institute. He is a Co-founder and Chair of Climate Change AI and serves as Scientific Co-director of Sustainability in the Digital Age. Dr. Rolnick received his Ph.D. in Applied Mathematics from MIT. He is a former NSF Mathematical Sciences Postdoctoral Research Fellow, NSF Graduate Research Fellow, and Fulbright Scholar, and was named to the MIT Technology Review’s 2021 list of “35 Innovators Under 35.”

Title of the talk:

Graph neural networks for catalyst evaluation

Abstract: Electrocatalysts will play a crucial role in the energy transition, from renewable energy storage to electrofuel synthesis. However, evaluation of candidate electrocatalysts using traditional methods is computationally expensive. Deep learning tools are increasingly being used to rapidly approximate physics-based simulations, but such algorithms continue to be limited in both accuracy and scalability. In this talk, we introduce a novel framework for graph neural networks to improve both computational efficiency and accuracy on the Open Catalyst 2020 (OC20) challenge and similar problems. We incorporate known atomic properties into the algorithm, while reducing the total amount of information that must be fed in as input. Furthermore, we introduce a flexible framework of data augmentation that enforces spatial symmetries across atomic configurations, while remaining computationally lightweight. We prove the validity of our method theoretically and demonstrate its superior performance on the OC20 dataset (S2EF, IS2RE) as well as common molecular modelling tasks (QM9, QM7-X).