Computational Science

Project Overview

We are working on applications of machine learning to computational science. We are particularly interested in the use of deep networks and differentiable programming for simulation of physical systems. One line of work uses deep networks to predict the evolution of physical systems at high fidelity. Another makes physics simulators differentiable, so that deep networks that act on or reason about physical systems can be trained end-to-end by backpropagating through physics.


Guaranteed Conservation of Momentum for Learning Particle-based Fluid Dynamics

Scale-invariant Learning by Physics Inversion

Differentiable Simulation of Soft Multi-body Systems

The h-index is no longer an effective correlate of scientific reputation

Efficient Differentiable Simulation of Articulated Bodies

Training Graph Neural Networks with 1000 Layers

Auto-decoding Graphs

Scalable Differentiable Physics for Learning and Control

Learning to Control PDEs with Differentiable Physics

Lagrangian Fluid Simulation with Continuous Convolutions

Differentiable Cloth Simulation for Inverse Problems