- Rating
- 3
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 3
Read review
Paper summary
This paper presents a novel graph foundation model for federated graph-level clustering, which aims to enable efficient knowledge transfer across diverse non-IID settings. The proposed framework learns transferable structural knowledge through adversarial pretraining, supports efficient local adaptation, and enables global knowledge sharing with only one round of communication. Extensive experiments demonstrate promising clustering performance and efficiency across different graph domains.
Critical strengths
The paper addresses an important problem in federated graph clustering by exploring how pretrained graph knowledge can be reused across different non-IID settings.
The method integrates structural information into graph pretraining and employs prototype alignment to facilitate knowledge sharing across heterogeneous clients.
The experiments cover 16 graph datasets and six non-IID settings, which demonstrates promising clustering performance. The ablation studies and generalization experiments further support the effectiveness of the proposed framework.
Critical weaknesses
The model is pretrained only on IMDB-BINARY, a relatively small social network dataset, yet it claims to generalize across diverse graph domains. Comparisons with standard graph pretraining methods and experiments using more diverse pretraining datasets would help substantiate this claim.
The proposed framework integrates several established techniques, including subgraph separation, adversarial graph generation, adapter tuning, and optimal transport-based prototype alignment. Although these components are organized into a unified framework, most of them follow existing methodological principles, and their combination does not clearly introduce a fundamentally new learning mechanism. The main contribution therefore appears to lie more in the integration of existing techniques than in substantial algorithmic innovation.
It remains unclear whether clusters from unrelated graph domains share meaningful semantics and how their alignment benefits local clustering.
The current objectives do not establish causal relationships, and the separation loss may allow trivial cluster assignments. The authors should clarify how such solutions are prevented.
Several supervised methods are adapted to unsupervised clustering, which may affect their original effectiveness. More controlled comparisons under comparable training and communication budgets would strengthen the evaluation.
Questions for authors
The model is pretrained only on IMDB-BINARY, a relatively small social network dataset, yet it claims to generalize across diverse graph domains. Comparisons with standard graph pretraining methods and experiments using more diverse pretraining datasets would help substantiate this claim.
The proposed framework integrates several established techniques, including subgraph separation, adversarial graph generation, adapter tuning, and optimal transport-based prototype alignment. Although these components are organized into a unified framework, most of them follow existing methodological principles, and their combination does not clearly introduce a fundamentally new learning mechanism. The main contribution therefore appears to lie more in the integration of existing techniques than in substantial algorithmic innovation.
It remains unclear whether clusters from unrelated graph domains share meaningful semantics and how their alignment benefits local clustering.
The current objectives do not establish causal relationships, and the separation loss may allow trivial cluster assignments. The authors should clarify how such solutions are prevented.
Several supervised methods are adapted to unsupervised clustering, which may affect their original effectiveness. More controlled comparisons under comparable training and communication budgets would strengthen the evaluation.
4 comments
Response to reviewer scmJ (2/3)
Response to W3&Q3:
Alignment is based on three quantities: the distribution of graph embeddings, the distribution of structural tokens, and the relationships among local prototypes (Eqs. 12–13). These descriptors are constructed using a shared pretrained backbone and structural vocabulary. Global prototypes consequently represent reference distributions in this common space, rather than predefined categories whose meanings must be identical across domains. Relational alignment supplements individual prototype similarity with information about how prototypes relate to one another. This shared information benefits local clustering through the feedback stage: clients first adapt to their own graphs, then use the aligned global references to refine their predictions. Unbalanced transport accommodates differences in prototype mass, while confidence-weighted feedback modulates the influence of the inferred correspondences.
The w/o OSC ablation in Fig. 3 shows that removing coordination reduces clustering performance in the evaluated settings. Table 7 further shows that nearest-neighbor matching and transport without the relational term () both underperform relational alignment, with the pretrained encoder and feedback procedure fixed. These results support the benefits of both cross-client sharing and relational information beyond individual prototype similarity. Inter-cluster relationships provide additional context for comparing heterogeneous prototypes and constructing global references for local prediction refinement. This transfer relies on partial compatibility of clustering-relevant distributional and structural patterns, without requiring identical class semantics across domains.
Response to W4&Q4:
We clarify that “causal subgraph” refers to a causally inspired design: the objective encourages the retained subgraph to preserve clustering information across structural perturbations while reducing its statistical dependence on the complementary component. These objectives do not establish causal identification. We have revised the terminology in the main text to emphasize the causally inspired nature of the separation. Regarding trivial solutions, the separation loss is optimized jointly with the contrastive and reconstruction objectives in Eq. (8), rather than in isolation. The edge-ratio penalty discourages all-edge and no-edge masks, while contrastive learning encourages discriminative graph representations and reconstruction preserves structural information. During local adaptation, the entropy terms in Eq. (11) further discourage uninformative cluster assignments. These mechanisms address different forms of degeneracy, although they do not formally exclude every constant-assignment solution during pretraining.
Response to reviewer scmJ (3/3)
Response to W5&Q5:
Thank you for the questions. To address the concern that removing supervision may disadvantage these methods, we additionally evaluated FedStar, CEFGL, FedGFM+, FedBook, and FedGALA using 2%, 5%, and 10% of the training labels with their original supervised objectives on four non-IID settings. These baselines use 10 communication rounds, while GFM-FGC uses no labels and one coordination round. As shown in Section 3.7 and Fig. 6, GFM-FGC achieves the best NMI and ARI across all four settings and label ratios. Its advantage therefore persists when the baselines receive training labels and additional communication opportunities. We also specify the training settings and cumulative costs in Appendices B.2 and D. The following are the evaluated model performance results.
SM-BIO non-IID setting
| Method | NMI:2% label | NMI:5% label | NMI:10% label | ARI:2% label | ARI:5% label | ARI:10% label |
|---|---|---|---|---|---|---|
| FedStar | 8.1 | 6.6 | 8.7 | 7.4 | 6.2 | 7.1 |
| CEFGL | 3.0 | 4.8 | 4.7 | 2.3 | 3.4 | 3.7 |
| FedGFM+ | 3.8 | 7.6 | 8.0 | 4.2 | 5.8 | 5.5 |
| FedBook | 5.8 | 7.3 | 6.8 | 3.5 | 5.2 | 6.0 |
| FedGALA | 8.7 | 8.5 | 9.2 | 6.4 | 7.6 | 6.8 |
SM-BIO-SY non-IID setting
| Method | NMI:2% label | NMI:5% label | NMI:10% label | ARI:2% label | ARI:5% label | ARI:10% label |
|---|---|---|---|---|---|---|
| FedStar | 6.5 | 9.8 | 10.2 | 7.1 | 10.8 | 11.8 |
| CEFGL | 3.4 | 3.7 | 3.6 | 3.4 | 3.9 | 3.9 |
| FedGFM+ | 8.3 | 7.8 | 8.9 | 8.6 | 10.3 | 9.3 |
| FedBook | 5.5 | 5.9 | 9.2 | 5.8 | 5.6 | 8.3 |
| FedGALA | 10.5 | 9.8 | 11.3 | 11.5 | 12.7 | 12.3 |
SN-SY non-IID setting
| Method | NMI:2% label | NMI:5% label | NMI:10% label | ARI:2% label | ARI:5% label | ARI:10% label |
|---|---|---|---|---|---|---|
| FedStar | 14.2 | 13.7 | 11.4 | 8.7 | 7.5 | 7.0 |
| CEFGL | 9.7 | 10.1 | 5.8 | 5.6 | 4.6 | 2.9 |
| FedGFM+ | 6.4 | 7.1 | 6.5 | 3.2 | 3.5 | 3.2 |
| FedBook | 9.0 | 8.8 | 10.3 | 4.4 | 5.5 | 5.1 |
| FedGALA | 12.0 | 8.0 | 13.8 | 6.6 | 4.2 | 7.9 |
CV non-IID setting
| Method | NMI:2% label | NMI:5% label | NMI:10% label | ARI:2% label | ARI:5% label | ARI:10% label |
|---|---|---|---|---|---|---|
| FedStar | 12.4 | 21.4 | 29.0 | 6.8 | 12.0 | 17.1 |
| CEFGL | 6.2 | 7.1 | 6.7 | 2.5 | 3.7 | 3.4 |
| FedGFM+ | 10.0 | 9.5 | 14.7 | 5.2 | 4.6 | 7.3 |
| FedBook | 9.0 | 18.5 | 25.7 | 4.3 | 9.8 | 14.7 |
| FedGALA | 31.8 | 21.9 | 34.2 | 17.8 | 9.2 | 20.5 |
Ours(0% label)
| Domain | NMI | ARI |
|---|---|---|
| SM-BIO | 11.8 | 10.0 |
| SM-BIO-SY | 14.8 | 16.9 |
| SN-SY | 18.7 | 11.9 |
| CV | 47.5 | 30.7 |
[1] You Y, Chen T, Sui Y, et al. Graph contrastive learning with augmentations[J]. Advances in neural information processing systems, 2020, 33: 5812-5823.
[2] Zheng Z, Yang Y, Guan Z, et al. Discrepancy-aware graph mask auto-encoder[C]//Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V. 2. 2025: 4038-4049.
[3] Bo J, Fang Y. CORE: Contrastive Masked Feature Reconstruction on Graphs[J]. arXiv preprint arXiv:2512.13235, 2025.
Thanks for the insightful comments again. We have provided a detailed response. We understand that Reviewer may have a busy schedule. Since only two days are left in the rebuttal period, we look forward to receiving your feedback and communication.
Response to reviewer scmJ (1/3)
We sincerely thank the reviewer for the questions raised, which have helped us improve the quality of our paper. We appreciate your recognition of the value contributed by this article. We also kindly request that the reviewer focus on our clarifications regarding these key issues. We have addressed the weaknesses and questions by labeling them W1–5 and Q1–5, respectively, and responding in that order. Please feel free to reach out if you have any further questions; we would be more than happy to answer them. We hope the response can address your concerns and lead to a more favorable assessment. The revised parts have been marked with blue underlines in the revised version.
Response to W1&Q1:
Thank you very much for the question. We have added comparisons with encoders pretrained using GraphCL [1], DGMAE [2], and CORE [3] on the same public data, using the same backbone and pretraining budget while keeping the local adaptation and sharing procedure fixed (Appendix C, Table 7). Section 3.6 and Fig. 5 further evaluate four alternative pretraining sources—uracil, Cuneiform, DBLP_v1, and TRIANGLES—across five downstream federated settings. The competitive performance across these sources supports the transfer capability of GFM-FGC beyond IMDB-BINARY pretraining.
Our pretraining outperforms GraphCL, DGMAE, and CORE, classic and advanced pretraining methods within the same adaptation and sharing pipeline. We attribute this advantage to its clustering-oriented design: subgraph separation encourages the encoder to preserve information supporting stable cluster assignments, while adversarial graph generation exposes it to additional structural variations. Together, these objectives encourage representations that retain clustering-relevant distinctions under structural variation, providing a more suitable initialization for local adaptation and subsequent prototype alignment.
In the revised version, we have included the above content in Appendix C.
Response to W2&Q2:
Thank you very much for your question; however, we respectfully disagree with the characterization of our contribution as merely combining existing techniques. We have incorporated them into an entirely new learning paradigm. Methodological innovation lies in formulating a new learning problem and developing the mechanisms needed to solve it. GFM-FGC addresses reusable, unsupervised federated graph-level clustering: knowledge is learned before target clients are observed, reused across different federated settings, and coordinated through a single communication round. This provides concrete advantages over the task-specific approaches considered in our comparison. The pretrained backbone and structural vocabulary are reused across deployments, avoiding repeated full-backbone training. Local adaptation optimizes lightweight modules, and one-shot coordination removes the dependence on iterative federated co-training. Moreover, this process requires neither ground-truth class labels nor language-model supervision. These capabilities arise from a specific connection between pretraining and deployment. Separation provides the structural reference for generation; generation expands the structural variations encountered during pretraining; and the same vocabulary subsequently describes client prototypes. Relational transport then coordinates independently adapted prototypes within the one-shot constraint. The contribution therefore lies in both the reusable learning paradigm and the seamlessly coupled mechanisms that make its deployment possible. Assessing novelty solely by whether each constituent technique already exists overlooks this methodological distinction.