1

Discovering Symmetry Groups with Flow Matching

Symmetry is fundamental to understanding physical systems and can improve performance and sample efficiency in machine learning. Both …

Yuxuan Chen, Jung Yeon Park, Floor Eijkelboom, Jianke Yang, Jan-Willem van de Meent, Lawson L.S. Wong, Robin Walters

Discovering Symmetry Groups with Flow Matching

Smoothness Errors in Dynamics Models and How to Avoid Them

Modern neural networks have shown promise for solving partial differential equations over surfaces, often by discretizing the surface …

Edward Berman, Luisa Li, Jung Yeon Park, Robin Walters

Smoothness Errors in Dynamics Models and How to Avoid Them

Approximate Equivariance in Reinforcement Learning

Equivariant neural networks have shown great success in reinforcement learning, improving sample efficiency and generalization when …

Jung Yeon Park, Sujay Bhatt, Sihan Zeng, Lawson L.S. Wong, Alec Koppel, Sumitra Ganesh, Robin Walters

Approximate Equivariance in Reinforcement Learning

Equivariant Action Sampling for Reinforcement Learning and Planning

Reinforcement learning (RL) algorithms for continuous control tasks require accurate sampling-based action selection. Many tasks, such …

Linfeng Zhao, Owen Lewis Howell, Xupeng Zhu, Jung Yeon Park, Zhewen Zhang, Robin Walters, Lawson L.S. Wong

Equivariant Action Sampling for Reinforcement Learning and Planning

Symmetric Models for Radar Response Modeling

Many radar applications require complex radar signature models that incorporate characteristics of an object’s shape and dynamics …

Colin Kohler, Nathan Vaska, Ramya Muthukrishnan, Whangbong Choi, Jung Yeon Park, Justin Goodwin, Rajmonda Caceres, Robin Walters

Symmetric Models for Radar Response Modeling

Modeling Dynamics over Meshes with Gauge Equivariant Nonlinear Message Passing

Data over non-Euclidean manifolds, often discretized as surface meshes, naturally arise in computer graphics and biological and …

Jung Yeon Park, Lawson L.S. Wong, Robin Walters

Modeling Dynamics over Meshes with Gauge Equivariant Nonlinear Message Passing

A General Theory of Correct, Incorrect, and Extrinsic Equivariance

Although equivariant machine learning has proven effective at many tasks, success depends heavily on the assumption that the ground …

Dian Wang, Xupeng Zhu, Jung Yeon Park, Robert Platt, Robin Walters

The Surprising Effectiveness of Equivariant Models in Domains with Latent Symmetry

Extensive work has demonstrated that equivariant neural networks can significantly improve sample efficiency and generalization by …

Dian Wang, Jung Yeon Park, Neel Sortur, Lawson L.S. Wong, Robin Walters, Robert Platt

The Surprising Effectiveness of Equivariant Models in Domains with Latent Symmetry

Robust Imitation Learning of a Few Demonstrations with a Backwards Model

Behavior cloning of expert demonstrations can speed up learning optimal policies in a more sample-efficient way over reinforcement …

Jung Yeon Park, Lawson L.S. Wong

Robust Imitation Learning of a Few Demonstrations with a Backwards Model

Learning Symmetric Embedding Networks for Equivariant World Models

Incorporating symmetries can lead to highly data-efficient and generalizable models by defining equivalence classes of data samples …

Jung Yeon Park, Ondrej Biza, Linfeng Zhao, Jan-Willem van de Meent, Robin Walters

Learning Symmetric Embedding Networks for Equivariant World Models