Saved in:
Bibliographic Details
Main Authors: Amboage, Juan, Röell, Ernst, Schnider, Patrick, Rieck, Bastian
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.00757
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918500436541440
author Amboage, Juan
Röell, Ernst
Schnider, Patrick
Rieck, Bastian
author_facet Amboage, Juan
Röell, Ernst
Schnider, Patrick
Rieck, Bastian
contents Graph neural networks (GNNs) largely rely on the message-passing paradigm, where nodes iteratively aggregate information from their neighbors. Yet, standard message passing neural networks (MPNNs) face well-documented theoretical and practical limitations. Graph positional encoding (PE) has emerged as a promising direction to address these limitations. The Euler Characteristic Transform (ECT) is an efficiently computable geometric-topological invariant that characterizes shapes and graphs. In this work, we combine the differentiable approximation of the ECT (DECT) and its local variant ($\ell$-ECT) to propose LEAP, a new end-to-end trainable local structural PE for graphs. We evaluate our approach on multiple real-world datasets as well as on a synthetic task designed to test its ability to extract topological features. Our results underline the potential of LEAP-based encodings as a powerful component for graph representation learning pipelines.
format Preprint
id arxiv_https___arxiv_org_abs_2510_00757
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle LEAP: Local ECT-Based Learnable Positional Encodings for Graphs
Amboage, Juan
Röell, Ernst
Schnider, Patrick
Rieck, Bastian
Machine Learning
Graph neural networks (GNNs) largely rely on the message-passing paradigm, where nodes iteratively aggregate information from their neighbors. Yet, standard message passing neural networks (MPNNs) face well-documented theoretical and practical limitations. Graph positional encoding (PE) has emerged as a promising direction to address these limitations. The Euler Characteristic Transform (ECT) is an efficiently computable geometric-topological invariant that characterizes shapes and graphs. In this work, we combine the differentiable approximation of the ECT (DECT) and its local variant ($\ell$-ECT) to propose LEAP, a new end-to-end trainable local structural PE for graphs. We evaluate our approach on multiple real-world datasets as well as on a synthetic task designed to test its ability to extract topological features. Our results underline the potential of LEAP-based encodings as a powerful component for graph representation learning pipelines.
title LEAP: Local ECT-Based Learnable Positional Encodings for Graphs
topic Machine Learning
url https://arxiv.org/abs/2510.00757