What Expressivity Theory Misses: Message Passing Complexity for GNNs

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kemper, Niklas, Wollschläger, Tom, Günnemann, Stephan
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917033520660480
author Kemper, Niklas
Wollschläger, Tom
Günnemann, Stephan
author_facet Kemper, Niklas
Wollschläger, Tom
Günnemann, Stephan
contents Expressivity theory, characterizing which graphs a GNN can distinguish, has become the predominant framework for analyzing GNNs, with new models striving for higher expressivity. However, we argue that this focus is misguided: First, higher expressivity is not necessary for most real-world tasks as these tasks rarely require expressivity beyond the basic WL test. Second, expressivity theory's binary characterization and idealized assumptions fail to reflect GNNs' practical capabilities. To overcome these limitations, we propose Message Passing Complexity (MPC): a continuous measure that quantifies the difficulty for a GNN architecture to solve a given task through message passing. MPC captures practical limitations like over-squashing while preserving the theoretical impossibility results from expressivity theory, effectively narrowing the gap between theory and practice. Through extensive validation on fundamental GNN tasks, we show that MPC's theoretical predictions correlate with empirical performance, successfully explaining architectural successes and failures. Thereby, MPC advances beyond expressivity theory to provide a more powerful and nuanced framework for understanding and improving GNN architectures.
format Preprint
id arxiv_https___arxiv_org_abs_2509_01254
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle What Expressivity Theory Misses: Message Passing Complexity for GNNs
Kemper, Niklas
Wollschläger, Tom
Günnemann, Stephan
Machine Learning
Expressivity theory, characterizing which graphs a GNN can distinguish, has become the predominant framework for analyzing GNNs, with new models striving for higher expressivity. However, we argue that this focus is misguided: First, higher expressivity is not necessary for most real-world tasks as these tasks rarely require expressivity beyond the basic WL test. Second, expressivity theory's binary characterization and idealized assumptions fail to reflect GNNs' practical capabilities. To overcome these limitations, we propose Message Passing Complexity (MPC): a continuous measure that quantifies the difficulty for a GNN architecture to solve a given task through message passing. MPC captures practical limitations like over-squashing while preserving the theoretical impossibility results from expressivity theory, effectively narrowing the gap between theory and practice. Through extensive validation on fundamental GNN tasks, we show that MPC's theoretical predictions correlate with empirical performance, successfully explaining architectural successes and failures. Thereby, MPC advances beyond expressivity theory to provide a more powerful and nuanced framework for understanding and improving GNN architectures.
title What Expressivity Theory Misses: Message Passing Complexity for GNNs
topic Machine Learning
url https://arxiv.org/abs/2509.01254