RAG-GFM: Overcoming In-Memory Bottlenecks in Graph Foundation Models via Retrieval-Augmented Generation

TL;DR

RAG-GFM introduces a retrieval-augmented approach to graph foundation models, externalizing knowledge to improve scalability, interpretability, and performance across various graph learning tasks.

Contribution

The paper proposes RAG-GFM, a novel retrieval-augmented framework that externalizes knowledge in GFMs using dual-modal retrieval and alignment, enhancing scalability and effectiveness.

Findings

Outperforms 13 state-of-the-art baselines on five benchmark datasets.

Achieves superior effectiveness and efficiency in node and graph classification.

Effectively externalizes knowledge, reducing in-memory bottlenecks.

Abstract

Graph Foundation Models (GFMs) have emerged as a frontier in graph learning, which are expected to deliver transferable representations across diverse tasks. However, GFMs remain constrained by in-memory bottlenecks: they attempt to encode knowledge into model parameters, which limits semantic capacity, introduces heavy lossy compression with conflicts, and entangles graph representation with the knowledge in ways that hinder efficient adaptation, undermining scalability and interpretability. In this work,we propose RAG-GFM, a Retrieval-Augmented Generation aided Graph Foundation Model that offloads knowledge from parameters and complements parameterized learning. To externalize graph knowledge, we build a dual-modal unified retrieval module, where a semantic store from prefix-structured text and a structural store from centrality-based motif. To preserve heterogeneous information, we design a dual-view alignment objective that contrasts both modalities to capture both content and relational patterns. To enable efficient downstream adaptation, we perform in-context augmentation to enrich supporting instances with retrieved texts and motifs as contextual evidence. Extensive experiments on five benchmark graph datasets demonstrate that RAG-GFM consistently outperforms 13 state-of-the-art baselines in both cross-domain node and graph classification, achieving superior effectiveness and efficiency.