Spectral Graph Neural Networks are Incomplete on Graphs with a Simple Spectrum

NeurIPSSpotlight2025

Authors
Snir Hordan, Maya Bechler-Speicher, Gur Lifshitz, Nadav Dym
Venue
NeurIPS 2025
Track
Spotlight

TL;DR

Spectral features are widely incorporated within Graph Neural Networks (GNNs) to improve their expressive power, or their ability to distinguish among non-isomorphic graphs…

Opening excerpt from the authors’ abstract. source

Read the paper

Topics

graph neural network graph

← All NeurIPS 2025 Spotlight papers · Browse the whole archive