arXiv:2510.26451v2 Announce Type: replace 
Abstract: Graph condensation (GC) has gained significant attention for its ability to synthesize smaller yet informative graphs. However, existing studies often overlook the robustness of GC in scenarios where the original graph is corrupted. In such cases, we observe that the performance of GC deteriorates significantly, while existing robust graph learning technologies offer only limited effectiveness. Through both empirical investigation and theoretical analysis, we reveal that GC is inherently an intrinsic-dimension-reducing process, synthesizing a condensed graph with lower classification complexity. Although this property is critical for effective GC performance, it remains highly vulnerable to adversarial perturbations. To tackle this vulnerability and improve GC robustness, we adopt the geometry perspective of graph data manifold and propose a novel Manifold-constrained Robust Graph Condensation framework named MRGC. Specifically, we introduce three graph data manifold learning modules that guide the condensed graph to lie within a smooth, low-dimensional manifold with minimal class ambiguity, thereby preserving the classification complexity reduction capability of GC and ensuring robust performance under universal adversarial attacks. Extensive experiments demonstrate the robustness of \ModelName\ across diverse attack scenarios.

لقد أصبحت تكثيف الرسوم البيانية (GC) طريقة مهمة لإنشاء رسوم بيانية أصغر وأكثر معلوماتية، لكن فعاليتها تتعرض للخطر عندما يكون الرسم البياني الأصلي تالفًا. تسلط الأبحاث الحديثة الضوء على أن تقنيات التعلم القوي للرسوم البيانية الحالية غير كافية لمعالجة هذه المشكلة، مما يؤدي إلى انخفاض ملحوظ في أداء GC في ظل الظروف العدائية.

La condensación de grafos (GC) ha surgido como un método significativo para crear grafos más pequeños e informativos, pero su efectividad se ve comprometida cuando el grafo original está corrupto. Investigaciones recientes destacan que las tecnologías de aprendizaje de grafos robustos existentes son insuficientes para abordar este problema, lo que lleva a una notable disminución en el rendimiento de GC bajo condiciones adversariales.

La condensation de graphes (GC) est devenue une méthode importante pour créer des graphes plus petits et informatifs, mais son efficacité est compromise lorsque le graphe original est corrompu. Des recherches récentes soulignent que les technologies d'apprentissage de graphes robustes existantes sont insuffisantes pour résoudre ce problème, entraînant une baisse notable des performances de GC dans des conditions adversariales.

Graph condensation (GC) has emerged as a significant method for creating smaller, informative graphs, but its effectiveness is compromised when the original graph is corrupted. Recent research highlights that existing robust graph learning technologies are insufficient in addressing this issue, leading to a notable decline in GC performance under adversarial conditions.

Robust Graph Condensation via Classification Complexity Mitigation

Was this article worth reading? Share it

LucidQuery AI

Graphix Calc

Graphite Note

ZeroGPT.org

GPTHumanizer

MGC Icon System

Ready to build your own newsroom?