A Learnable Multigrid Framework via Graph Convolutions
Abstract
This paper presents a novel framework that integrates learnable graph convolutions with the geometric multigrid method for solving partial differential equations (PDEs). The discretization of PDEs is first represented as a graph structure, enabling the application of graph convolutions to enhance the multigrid performance. By incorporating graph convolutions into the multigrid components such as smoothing and inter-grid transfer operators, we develop a learnable multigrid that can adaptively optimize its performance based on the underlying problem characteristics. In this framework, the graph convolutions are embedded directly within the multigrid cycle, effectively transforming the entire multigrid solver into a specialized neural network architecture, rather than combining a classical solver with surrogate models. The learnable multigrid framework is lightweight in terms of parameter count and requires minimal training effort to achieve good performance. Numerical experiments demonstrate the effectiveness of the proposed approach in solving some challenging problems, showing improved convergence rates compared to traditional multigrid methods. The generalizability of the learned parameters across different problem settings, including varying source terms, coefficients, geometries, and mesh sizes, is also investigated with proper weight-sharing and transfer-learning strategies.
Related
- Actually, All Learnable Distribution Classes Are Privately Learnable
- A Reduced Coupling-Space Multigrid Schur Preconditioner for Fictitious-Domain Interface Problems
- Shifted convolutions of divisor functions
- Robust Multigrid on Adaptive Meshes for the Integral Fractional Laplacian
- Data-driven accelerated multigrid method for Riesz space fractional diffusion equations
BibTeX
@misc{wang2026learnable,
title = {{A Learnable Multigrid Framework via Graph Convolutions}},
author = {Xiuping Wang and Li Luo},
howpublished = {arXiv},
year = {2026},
month = aug,
doi = {10.48550/arxiv.2608.29082},
eprint = {2608.29082},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2608.29082},
}
RIS
TY - UNPB
TI - A Learnable Multigrid Framework via Graph Convolutions
AU - Xiuping Wang
AU - Li Luo
PY - 2026
DA - 2026/08/29
PB - arXiv
DO - 10.48550/arxiv.2608.29082
UR - https://arxiv.org/abs/2608.29082
AB - This paper presents a novel framework that integrates learnable graph convolutions with the geometric multigrid method for solving partial differential equations (PDEs). The discretization of PDEs is first represented as a graph structure, enabling the application of graph convolutions to enhance the multigrid performance. By incorporating graph convolutions into the multigrid components such as smoothing and inter-grid transfer operators, we develop a learnable multigrid that can adaptively optimize its performance based on the underlying problem characteristics. In this framework, the graph convolutions are embedded directly within the multigrid cycle, effectively transforming the entire multigrid solver into a specialized neural network architecture, rather than combining a classical solver with surrogate models. The learnable multigrid framework is lightweight in terms of parameter count and requires minimal training effort to achieve good performance. Numerical experiments demonstrate the effectiveness of the proposed approach in solving some challenging problems, showing improved convergence rates compared to traditional multigrid methods. The generalizability of the learned parameters across different problem settings, including varying source terms, coefficients, geometries, and mesh sizes, is also investigated with proper weight-sharing and transfer-learning strategies.
ER -