Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Multi-Domain Riemannian Graph Gluing for Building Graph Foundation Models

About

Multi-domain graph pre-training integrates knowledge from diverse domains to enhance performance in the target domains, which is crucial for building graph foundation models. Despite initial success, existing solutions often fall short of answering a fundamental question: how is knowledge integrated or transferred across domains? This theoretical limitation motivates us to rethink the consistency and transferability between model pre-training and domain adaptation. In this paper, we propose a fresh Riemannian geometry perspective, whose core idea is to merge any graph dataset into a unified, smooth Riemannian manifold, enabling a systematic understanding of knowledge integration and transfer. To achieve this, our key contribution is the theoretical establishment of neural manifold gluing, which first characterizes local geometry using an adaptive orthogonal frame and then "glues" the local pieces together into a coherent whole. Building on this theory, we present the GraphGlue framework, which supports batched pre-training with EMA prototyping and provides a transferability measure based on geometric consistence. Extensive experiments demonstrate its superior performance across diverse graph domains. Moreover, we empirically validated GraphGlue's geometric scaling law, showing that larger quantities of datasets improve model transferability by producing a smoother manifold. Codes are available at https://github.com/RiemannGraph/GraphGlue.

Li Sun, Zhenhao Huang, Silei Chen, Lanxu Yang, Junda Ye, Sen Su, Philip S. Yu• 2026

Related benchmarks

TaskDatasetResultRank
Graph ClassificationPROTEINS
Accuracy69.74
1252
Node ClassificationWisconsin
Accuracy51.49
864
Node ClassificationTexas
Accuracy0.526
801
Node ClassificationRoman-Empire
Accuracy20.67
327
Node Classificationamazon-ratings
Accuracy36.26
309
Node ClassificationarXiv
Accuracy39.98
254
Node ClassificationREDDIT
Accuracy85.05
216
Node ClassificationComputers
Mean Accuracy73.29
169
Node ClassificationComputers
Accuracy74.15
145
Link ClassificationFB15k-237
Accuracy81.51
97
Showing 10 of 19 rows

Other info

GitHub

Follow for update