A Cross-graph Tuning-free GNN Prompting Framework

TL;DR

This paper introduces CTP, a tuning-free GNN prompting framework that generalizes across different graphs without retraining, significantly improving few-shot prediction accuracy.

Contribution

The proposed CTP framework enables tuning-free, plug-and-play GNN inference across diverse graphs, addressing limitations of existing prompt methods.

Findings

Achieves an average accuracy gain of 30.8% over SOTA methods.

Demonstrates effectiveness on both homogeneous and heterogeneous graphs.

Supports deployment on unseen graphs without parameter tuning.

Abstract

GNN prompting aims to adapt models across tasks and graphs without requiring extensive retraining. However, most existing graph prompt methods still require task-specific parameter updates and face the issue of generalizing across graphs, limiting their performance and undermining the core promise of prompting. In this work, we introduce a Cross-graph Tuning-free Prompting Framework (CTP), which supports both homogeneous and heterogeneous graphs, can be directly deployed to unseen graphs without further parameter tuning, and thus enables a plug-and-play GNN inference engine. Extensive experiments on few-shot prediction tasks show that, compared to SOTAs, CTP achieves an average accuracy gain of 30.8% and a maximum gain of 54%, confirming its effectiveness and offering a new perspective on graph prompt learning.