Advancing Graph Few-Shot Learning via In-Context Learning

2026-05-23

AI Summary

This work addresses the challenges in graph few-shot learning where unlabeled nodes are underutilized and inference typically requires complex fine-tuning. To overcome these limitations, the authors propose a fine-tuning-free sequential inference framework that leverages in-context learning through an unsupervised meta-learning paradigm coupled with a task-adaptive pseudo-task generation mechanism. Within a single forward pass, the model dynamically constructs class-aware representations for query nodes. Key innovations include role-based embedding initialization, a dual-context module integrating local topological structure and global task dependencies, and a structure-adaptive pseudo-task generator informed by unlabeled data. Extensive experiments demonstrate that the proposed method significantly outperforms existing approaches across multiple benchmark datasets, achieving both high efficiency and superior performance in few-shot node classification.

Abstract

Graph few-shot learning, which aims to classify nodes from novel classes with only a few labeled examples, is a widely studied problem in graph learning. However, existing methods often face two key limitations. First, the predominant graph few-shot learning paradigm relies on supervised tasks, failing to leverage the vast number of unlabeled nodes in the graph. Second, many approaches require complex task adaptation or fine-tuning during inference, limiting their efficiency and applicability. Inspired by the powerful in-context learning capabilities of large language models, we propose a novel model named VISION for adVancIng graph few-Shot learning via In-cOntext LearNing to address these challenges. Our model reframes graph few-shot learning as a fine-tuning-free sequence reasoning problem. At its core is a context-aware network that initializes nodes with role embeddings and employs a dual-context fusion module to synergistically integrate local topological structures and global task-level dependencies. This allows our model to dynamically generate class-aware representations for the query set conditioned on the support set context in a single forward pass. To effectively train our model, we introduce an unsupervised task generator that creates structure-adaptive features and constructs diverse pseudo-tasks from abundant unlabeled data. Our method unifies unsupervised meta-learning with graph in-context learning, achieving efficient inference. Extensive experiments on multiple benchmark datasets demonstrate the superiority of our model. Our public code can be found

Problem

• Graph Few-Shot Learning
• In-Context Learning
• Unlabeled Nodes
• Task Adaptation
• Efficient Inference

Innovation

• in-context learning
• graph few-shot learning
• unsupervised meta-learning
• dual-context fusion
• role embeddings