MAGPrompt: Message-Adaptive Graph Prompt Tuning for Graph Neural Networks

TL;DR

This paper introduces message-adaptive graph prompt tuning, a method that enhances pre-trained GNNs by injecting learnable prompts into message passing, improving adaptation to downstream tasks efficiently.

Contribution

It proposes a novel prompt tuning approach that modifies message passing in GNNs, enabling better task adaptation without retraining the entire model.

Findings

Consistent performance improvements in few-shot learning scenarios.

Competitive results with fine-tuning in full-shot settings.

Compatibility with various GNN architectures and pre-training methods.

Abstract

Pre-trained graph neural networks (GNNs) transfer well, but adapting them to downstream tasks remains challenging due to mismatches between pre-training objectives and task requirements. Graph prompt tuning offers a parameter-efficient alternative to fine-tuning, yet most methods only modify inputs or representations and leave message passing unchanged, limiting their ability to adapt neighborhood interactions. We propose message-adaptive graph prompt tuning, which injects learnable prompts into the message passing step to reweight incoming neighbor messages and add task-specific prompt vectors during message aggregation, while keeping the backbone GNN frozen. The approach is compatible with common GNN backbones and pre-training strategies, and applicable across downstream settings. Experiments on diverse node- and graph-level datasets show consistent gains over prior graph prompting methods in few-shot settings, while achieving performance competitive with fine-tuning in full-shot regimes.