H4G: Unlocking Faithful Inference for Zero-Shot Graph Learning in Hyperbolic Space
Open Access
- 8 August 2026
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 1851-1861
Abstract
Text-attributed graphs are widely used across domains, offering rich opportunities for zero-shot learning via graph-text alignment. However, existing methods struggle with tasks requiring fine-grained pattern recognition, particularly on heterophilic graphs. Through empirical and theoretical analysis, we identify an over-abstraction problem: current approaches operate at excessively large hyperbolic radii, compressing multi-scale structural information into uniform high-level abstractions. This abstraction-induced information loss obscures critical local patterns essential for accurate predictions. By analyzing embeddings in hyperbolic space, we demonstrate that optimal graph learning requires faithful preservation of fine-grained structural details, better retained by representations positioned closer to the origin. To address this, we propose H4G, a framework that systematically reduces embedding radii using learnable block-diagonal scaling matrices and Möbius matrix multiplication. This approach restores access to fine-grained patterns while maintaining global receptive ability with minimal computational overhead. Experiments show H4G achieves state-of-the-art zero-shot performance with 12.8% improvement on heterophilic graphs and 8.4% on homophilic graphs, confirming that radius reduction enables faithful multi-scale representation for advancing zero-shot graph learning.Keywords
This publication has 25 references indexed in Scilit:
- Hyperbolic Bernstein Neural Networks: Enhancing graph convolutions in non-Euclidean spacesNeural Networks, 2025
- DHHNN: A Dynamic Hypergraph Hyperbolic Neural Network based on variational autoencoder for multimodal data integration and node classificationInformation Fusion, 2025
- Self-Cumulative Contrastive Graph ClusteringIEEE/CAA Journal of Automatica Sinica, 2025
- Self-supervised graph contrastive learning with diffusion augmentation for functional MRI analysis and brain disorder detectionMedical Image Analysis, 2024
- Data‐efficient graph learning: Problems, progress, and prospectsAI Magazine, 2024
- Generalized Few-Shot Node Classification With Graph Knowledge DistillationIEEE Transactions on Computational Social Systems, 2024
- Graph Structured Neural Networks for Perturbation BiologybioRxiv, 2024
- Graph Prototypical Networks for Few-shot Learning on Attributed Networks2020
- Cluster-GCN2019
- Image-Based Recommendations on Styles and Substitutes2015