UFO: A Unified Flow-Oriented Framework for Robust Continual Graph Learning

2026-05-10

AI Summary

This work addresses the co-occurrence of catastrophic forgetting and label noise in continual graph learning, revealing and formally defining a novel phenomenon termed “catastrophic memorization” induced by noisy supervision. To tackle this challenge, the authors propose UFO, a unified flow-oriented framework that leverages streaming generative modeling to learn conditional feature distributions, thereby enabling replay-free continual learning. UFO further incorporates an instance-level reliability scoring mechanism to effectively distinguish clean from noisy nodes, jointly mitigating both catastrophic forgetting and noise-induced interference. Extensive experiments on four graph benchmark datasets demonstrate that UFO consistently outperforms existing methods across varying noise ratios, achieving state-of-the-art performance in both accuracy and forgetting metrics.

Abstract

Graph learning research has increasingly shifted toward continual graph learning (CGL), which better reflects real-world scenarios where graphs evolve over time. However, existing CGL methods largely assume clean supervision and overlook a critical challenge: the newly arriving portions of the graph are often noisy, due to annotation errors or adversarial corruption. This mismatch limits their applicability in practice. In this work, we study robust continual graph learning, where models must simultaneously handle catastrophic forgetting and noisy supervision in evolving graph data. We show that label noise introduces a new failure mode, catastrophic remembering, where models persistently reinforce corrupted knowledge across tasks. To address these challenges, we propose a Unified Flow-Oriented framework (UFO). First, UFO models conditional feature distributions via flow-based generative modeling and produces replay representations, mitigating forgetting without storing historical data. Second, UFO estimates instance-level reliability scores to distinguish clean from noisy nodes, reducing the impact of corrupted supervision and alleviating catastrophic remembering. Extensive experiments on four benchmark graph datasets under varying noise ratios demonstrate that UFO consistently outperforms existing methods in both accuracy and forgetting metrics. Code is available at: https://anonymous.4open.science/r/UFO.

Problem

• continual graph learning
• label noise
• catastrophic forgetting
• catastrophic remembering
• noisy supervision

Innovation

• continual graph learning
• flow-based generative modeling
• catastrophic remembering
• noisy supervision
• replay-free