A Latent space solver for PDE generalization

no code implementations • 6 Apr 2021 • Rishikesh Ranade, Chris Hill, Haiyang He, Amir Maleki, Jay Pathak

In this work we propose a hybrid solver to solve partial differential equation (PDE)s in the latent space.

Geometry encoding for numerical simulations

1 code implementation • ICLR Workshop GTRL 2021 • Amir Maleki, Jan Heyse, Rishikesh Ranade, Haiyang He, Priya Kasimbeg, Jay Pathak

We present a notion of geometry encoding suitable for machine learning-based numerical simulation.

BIG-bench Machine Learning

OVERT: An Algorithm for Safety Verification of Neural Network Control Policies for Nonlinear Systems

2 code implementations • 3 Aug 2021 • Chelsea Sidrane, Amir Maleki, Ahmed Irfan, Mykel J. Kochenderfer

In response to this challenge, we present OVERT: a sound algorithm for safety verification of nonlinear discrete-time closed loop dynamical systems with neural network control policies.

A composable autoencoder-based algorithm for accelerating numerical simulations

no code implementations • 29 Sep 2021 • Rishikesh Ranade, Derek Christopher Hill, Haiyang He, Amir Maleki, Norman Chang, Jay Pathak

Numerical simulations for engineering applications solve partial differential equations (PDE) to model various physical processes.

A composable autoencoder-based iterative algorithm for accelerating numerical simulations

no code implementations • 7 Oct 2021 • Rishikesh Ranade, Chris Hill, Haiyang He, Amir Maleki, Norman Chang, Jay Pathak

Numerical simulations for engineering applications solve partial differential equations (PDE) to model various physical processes.

BIG-bench Machine Learning