Multi-Domain Graph Foundation Models: Robust Knowledge Transfer via Topology Alignment

Shuo Wang, Bokui Wang, Zhixiang Shen, Boyan Deng, Zhao Kang
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:64806-64821, 2025.

Abstract

Recent advances in CV and NLP have inspired researchers to develop general-purpose graph foundation models through pre-training across diverse domains. However, a fundamental challenge arises from the substantial differences in graph topologies across domains. Additionally, real-world graphs are often sparse and prone to noisy connections and adversarial attacks. To address these issues, we propose the Multi-Domain Graph Foundation Model (MDGFM), a unified framework that aligns and leverages cross-domain topological information to facilitate robust knowledge transfer. MDGFM bridges different domains by adaptively balancing features and topology while refining original graphs to eliminate noise and align topological structures. To further enhance knowledge transfer, we introduce an efficient prompt-tuning approach. By aligning topologies, MDGFM not only improves multi-domain pre-training but also enables robust knowledge transfer to unseen domains. Theoretical analyses provide guarantees of MDGFM’s effectiveness and domain generalization capabilities. Extensive experiments on both homophilic and heterophilic graph datasets validate the robustness and efficacy of our method.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-wang25dj, title = {Multi-Domain Graph Foundation Models: Robust Knowledge Transfer via Topology Alignment}, author = {Wang, Shuo and Wang, Bokui and Shen, Zhixiang and Deng, Boyan and Kang, Zhao}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {64806--64821}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/wang25dj/wang25dj.pdf}, url = {https://proceedings.mlr.press/v267/wang25dj.html}, abstract = {Recent advances in CV and NLP have inspired researchers to develop general-purpose graph foundation models through pre-training across diverse domains. However, a fundamental challenge arises from the substantial differences in graph topologies across domains. Additionally, real-world graphs are often sparse and prone to noisy connections and adversarial attacks. To address these issues, we propose the Multi-Domain Graph Foundation Model (MDGFM), a unified framework that aligns and leverages cross-domain topological information to facilitate robust knowledge transfer. MDGFM bridges different domains by adaptively balancing features and topology while refining original graphs to eliminate noise and align topological structures. To further enhance knowledge transfer, we introduce an efficient prompt-tuning approach. By aligning topologies, MDGFM not only improves multi-domain pre-training but also enables robust knowledge transfer to unseen domains. Theoretical analyses provide guarantees of MDGFM’s effectiveness and domain generalization capabilities. Extensive experiments on both homophilic and heterophilic graph datasets validate the robustness and efficacy of our method.} }
Endnote
%0 Conference Paper %T Multi-Domain Graph Foundation Models: Robust Knowledge Transfer via Topology Alignment %A Shuo Wang %A Bokui Wang %A Zhixiang Shen %A Boyan Deng %A Zhao Kang %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-wang25dj %I PMLR %P 64806--64821 %U https://proceedings.mlr.press/v267/wang25dj.html %V 267 %X Recent advances in CV and NLP have inspired researchers to develop general-purpose graph foundation models through pre-training across diverse domains. However, a fundamental challenge arises from the substantial differences in graph topologies across domains. Additionally, real-world graphs are often sparse and prone to noisy connections and adversarial attacks. To address these issues, we propose the Multi-Domain Graph Foundation Model (MDGFM), a unified framework that aligns and leverages cross-domain topological information to facilitate robust knowledge transfer. MDGFM bridges different domains by adaptively balancing features and topology while refining original graphs to eliminate noise and align topological structures. To further enhance knowledge transfer, we introduce an efficient prompt-tuning approach. By aligning topologies, MDGFM not only improves multi-domain pre-training but also enables robust knowledge transfer to unseen domains. Theoretical analyses provide guarantees of MDGFM’s effectiveness and domain generalization capabilities. Extensive experiments on both homophilic and heterophilic graph datasets validate the robustness and efficacy of our method.
APA
Wang, S., Wang, B., Shen, Z., Deng, B. & Kang, Z.. (2025). Multi-Domain Graph Foundation Models: Robust Knowledge Transfer via Topology Alignment. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:64806-64821 Available from https://proceedings.mlr.press/v267/wang25dj.html.

Related Material