Graph Contrastive Learning (GCL) relies on semantically consistent graph augmentations, but common local perturbations provide limited control over global structural consistency, motivating a more principled global augmentation strategy. We therefore propose Fractal Graph Contrastive Learning (FractalGCL), a theory-motivated framework that constructs a renormalisation-based augmented graph and introduces a fractal-dimension-aware contrastive loss that penalises unreliable positive views and reweights negative-pair repulsion by finite-scale box-counting discrepancies. However, computing these discrepancies introduces substantial overhead, so we derive and justify a Gaussian surrogate that avoids repeated box-counting on renormalised graphs, yielding about a 61% runtime reduction. Experiments show that FractalGCL serves as an effective frozen-pretraining tool on MalNet-Tiny, achieves strong performance on the standard TUDataset benchmarks, and outperforms the next-best method on real-world urban traffic tasks by 4.51 percentage points in average accuracy. Code is available at https://anonymous.4open.science/r/FractalGCL-0511/.
@article{arxiv.2505.11356,
title = {Fractal Graph Contrastive Learning},
author = {Nero Z. Li and Xuehao Zhai and Zhichao Shi and Boshen Shi and Xuhui Jiang},
journal= {arXiv preprint arXiv:2505.11356},
year = {2026}
}