MAGPrompt: 用于图神经网络的消息适应图提示调优
MAGPrompt: Message-Adaptive Graph Prompt Tuning for Graph Neural Networks
AI总结:
MAGPrompt通过在消息传递步骤中注入可学习提示,实现对图神经网络的高效适应,在少样本设置中优于现有方法,并在全样本设置中与微调性能相当。
摘要:
预训练的图神经网络(GNN)在迁移学习中表现良好,但适应下游任务仍然具有挑战性,因为预训练目标与任务需求之间存在不匹配。图提示调优提供了一种参数高效的替代方法,但大多数方法只修改输入或表示,而保留消息传递不变,限制了它们适应邻域交互的能力。我们提出了消息适应的图提示调优,该方法在消息传递步骤中注入可学习的提示,以重新加权传入的邻居消息,并在消息聚合过程中添加任务特定的提示向量,同时保持基础GNN冻结。该方法与常见的GNN基础架构和预训练策略兼容,并适用于各种下游设置。在多样化的节点级和图级数据集上的实验表明,在少样本设置中相对于先前的图提示方法有持续的提升,同时在全样本设置中达到与微调相当的性能。
英文摘要:
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.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。