---
title: "Beyond Homophily: Spectrum-Based Graph Pre-Training and Cluster-Augmented Prompt Tuning"
authors:
  - "Tingting Li"
  - "Yonghao Li"
  - "Xiangkun Wang"
  - "Sixiang Chen"
  - "Boyang Fan"
  - "Lingfei Ren"
  - "Hao Yu"
  - "Xin Yang"
doi: "10.24963/ijcai.2026/310"
url: "https://ijcai.org/proceedings/2026/310"
pdf: "https://ijcai.org/proceedings/2026/0310.pdf"
venue: "Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence"
year: 2026
pages: "2789-2797"
track: "Main Track"
keywords:
  - "Data Mining: Mining graphs"
  - "Data Mining: Networks"
  - "Machine Learning: Few-shot learning"
---

# Beyond Homophily: Spectrum-Based Graph Pre-Training and Cluster-Augmented Prompt Tuning

Tingting Li, Yonghao Li, Xiangkun Wang, Sixiang Chen, Boyang Fan, Lingfei Ren, Hao Yu, Xin Yang

*Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence*, pages 2789-2797. DOI: [10.24963/ijcai.2026/310](https://doi.org/10.24963/ijcai.2026/310)

## Abstract

Graph pre-training and prompt tuning provide an effective route to label-efficient node classification by learning transferable backbones and adapting them with lightweight prompts. However, existing pre-train-and-prompt pipelines often generalize poorly across graphs with diverse homophily due to two fundamental limitations: (i) spectral bias, where learned representations overemphasize a single frequency band, and (ii) prompt misalignment, where spatial-domain prompts cannot explicitly modulate critical spectral components and may even disrupt frequency-aware backbones. In this paper, we propose SCA-GPPT, a spectrum-aware framework that unifies Spectrum-based Graph Pre-training and Cluster-Augmented Prompt Tuning. First, we present a systematic spectral analysis of graph prompting, revealing the inherent limitations of spatial-domain prompts in capturing attenuated high-frequency signals. Second, we explicitly learn complementary low- and high-frequency spectral filters via parameterized Chebyshev polynomials and employ band-aware contrastive learning to model diverse spectral patterns during pre-training. For downstream adaptation, we freeze the pre-trained backbone and introduce cluster-augmented prompts that directly adapt Chebyshev coefficients while enforcing global structural consistency, enabling stable few-shot transfer across varying homophily regimes. Extensive experiments on real-world benchmarks demonstrate that SCA-GPPT consistently outperforms strong baselines under both transductive and inductive settings. Our code is available at https://anonymous.4open.science/r/SCA-GPPT-17B2.