Saved in:
Bibliographic Details
Main Authors: Wang, Xiaohan, Bo, Deyu, Li, Longlong, Xia, Kelin
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.05759
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917527621206016
author Wang, Xiaohan
Bo, Deyu
Li, Longlong
Xia, Kelin
author_facet Wang, Xiaohan
Bo, Deyu
Li, Longlong
Xia, Kelin
contents It is well established that spectral graph neural networks (GNNs) can universally approximate node signals; however, their expressive power remains bounded by the 1-dimensional Weisfeiler-Lehman test, which is mirrored in their lack of universality for higher-order signals. To go beyond this bound, we propose the Full-Spectrum GNNs (FSpecGNNs), a second-order generalization of classical spectral GNNs. FSpecGNN advances spectral filtering from two perspectives: (1) it lifts signals from the node domain to the node-pair domain; and (2) it extends the univariate spectral filter over eigenvalues to a bivariate filter over eigenvalue pairs. We show that classical spectral GNNs arise as a diagonal special case of FSpecGNNs, and prove that FSpecGNNs can be at most as expressive as Local 2-GNN while universally approximating node-pair signals, the latter being particularly beneficial for heterophilic graph learning. Moreover, FSpecGNN admits scalable implementations that avoid explicit node-pair-level computations; combined with a low-rank approximation that reduces full-spectrum convolution to a combination of polynomial spectral filters, it enables learning on large graphs. Empirically, FSpecGNN validates the predicted expressivity and delivers strong performance on heterophilic benchmarks.
format Preprint
id arxiv_https___arxiv_org_abs_2605_05759
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Full-Spectrum Graph Neural Networks: Expressive and Scalable
Wang, Xiaohan
Bo, Deyu
Li, Longlong
Xia, Kelin
Machine Learning
It is well established that spectral graph neural networks (GNNs) can universally approximate node signals; however, their expressive power remains bounded by the 1-dimensional Weisfeiler-Lehman test, which is mirrored in their lack of universality for higher-order signals. To go beyond this bound, we propose the Full-Spectrum GNNs (FSpecGNNs), a second-order generalization of classical spectral GNNs. FSpecGNN advances spectral filtering from two perspectives: (1) it lifts signals from the node domain to the node-pair domain; and (2) it extends the univariate spectral filter over eigenvalues to a bivariate filter over eigenvalue pairs. We show that classical spectral GNNs arise as a diagonal special case of FSpecGNNs, and prove that FSpecGNNs can be at most as expressive as Local 2-GNN while universally approximating node-pair signals, the latter being particularly beneficial for heterophilic graph learning. Moreover, FSpecGNN admits scalable implementations that avoid explicit node-pair-level computations; combined with a low-rank approximation that reduces full-spectrum convolution to a combination of polynomial spectral filters, it enables learning on large graphs. Empirically, FSpecGNN validates the predicted expressivity and delivers strong performance on heterophilic benchmarks.
title Full-Spectrum Graph Neural Networks: Expressive and Scalable
topic Machine Learning
url https://arxiv.org/abs/2605.05759