From Basis to Basis: Gaussian Particle Representation for Interpretable PDE Operators

TL;DR

This paper introduces a Gaussian basis representation for PDE operators that enhances interpretability, handles high-frequency structures efficiently, and achieves near-linear complexity, outperforming existing neural operator methods.

Contribution

It proposes a novel Gaussian particle basis for PDE modeling, enabling explicit geometric interpretation and scalable, resolution-agnostic operator learning.

Findings

Achieves state-of-the-art accuracy on PDE benchmarks.

Provides intrinsic interpretability through explicit geometric parameters.

Supports irregular geometries and 2D-to-3D extension seamlessly.

Abstract

Learning PDE dynamics for fluids increasingly relies on neural operators and Transformer-based models, yet these approaches often lack interpretability and struggle with localized, high-frequency structures while incurring quadratic cost in spatial samples. We propose representing fields with a Gaussian basis, where learned atoms carry explicit geometry (centers, anisotropic scales, weights) and form a compact, mesh-agnostic, directly visualizable state. Building on this representation, we introduce a Gaussian Particle Operator that acts in modal space: learned Gaussian modal windows perform a Petrov-Galerkin measurement, and PG Gaussian Attention enables global cross-scale coupling. This basis-to-basis design is resolution-agnostic and achieves near-linear complexity in N for a fixed modal budget, supporting irregular geometries and seamless 2D-to-3D extension. On standard PDE benchmarks and real datasets, our method attains state-of-the-art competitive accuracy while providing intrinsic interpretability.