Saved in:
Bibliographic Details
Main Authors: Franks, Billy Joe, Eliasof, Moshe, Cantürk, Semih, Wolf, Guy, Schönlieb, Carola-Bibiane, Fellenz, Sophie, Kloft, Marius
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2412.07407
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912254553751552
author Franks, Billy Joe
Eliasof, Moshe
Cantürk, Semih
Wolf, Guy
Schönlieb, Carola-Bibiane
Fellenz, Sophie
Kloft, Marius
author_facet Franks, Billy Joe
Eliasof, Moshe
Cantürk, Semih
Wolf, Guy
Schönlieb, Carola-Bibiane
Fellenz, Sophie
Kloft, Marius
contents Recent advances in integrating positional and structural encodings (PSEs) into graph neural networks (GNNs) have significantly enhanced their performance across various graph learning tasks. However, the general applicability of these encodings and their potential to serve as foundational representations for graphs remain uncertain. This paper investigates the fine-tuning efficiency, scalability with sample size, and generalization capability of learnable PSEs across diverse graph datasets. Specifically, we evaluate their potential as universal pre-trained models that can be easily adapted to new tasks with minimal fine-tuning and limited data. Furthermore, we assess the expressivity of the learned representations, particularly, when used to augment downstream GNNs. We demonstrate through extensive benchmarking and empirical analysis that PSEs generally enhance downstream models. However, some datasets may require specific PSE-augmentations to achieve optimal performance. Nevertheless, our findings highlight their significant potential to become integral components of future graph foundation models. We provide new insights into the strengths and limitations of PSEs, contributing to the broader discourse on foundation models in graph learning.
format Preprint
id arxiv_https___arxiv_org_abs_2412_07407
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Towards Graph Foundation Models: A Study on the Generalization of Positional and Structural Encodings
Franks, Billy Joe
Eliasof, Moshe
Cantürk, Semih
Wolf, Guy
Schönlieb, Carola-Bibiane
Fellenz, Sophie
Kloft, Marius
Machine Learning
Recent advances in integrating positional and structural encodings (PSEs) into graph neural networks (GNNs) have significantly enhanced their performance across various graph learning tasks. However, the general applicability of these encodings and their potential to serve as foundational representations for graphs remain uncertain. This paper investigates the fine-tuning efficiency, scalability with sample size, and generalization capability of learnable PSEs across diverse graph datasets. Specifically, we evaluate their potential as universal pre-trained models that can be easily adapted to new tasks with minimal fine-tuning and limited data. Furthermore, we assess the expressivity of the learned representations, particularly, when used to augment downstream GNNs. We demonstrate through extensive benchmarking and empirical analysis that PSEs generally enhance downstream models. However, some datasets may require specific PSE-augmentations to achieve optimal performance. Nevertheless, our findings highlight their significant potential to become integral components of future graph foundation models. We provide new insights into the strengths and limitations of PSEs, contributing to the broader discourse on foundation models in graph learning.
title Towards Graph Foundation Models: A Study on the Generalization of Positional and Structural Encodings
topic Machine Learning
url https://arxiv.org/abs/2412.07407