Rethinking Client-oriented Federated Graph Learning

Guoren Wang; Rong-Hua Li; Xunkai Li; Yinlin Zhu; Zekai Chen

REVIEW 3 major objections 6 minor 34 references

Rethinking Client-oriented Federated Graph Learning

T0 review · 3 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read FedC4 sends each client a small distilled graph instead of raw node embeddings, and reports better accuracy and up to 1000x lower communication than prior federated graph learning methods.

desk verdict FedC4 has real empirical substance integrating graph condensation into client-to-client FGL, but the privacy proof in Sec 4.1 is a non-sequitur and the communication-cost analysis is under-derived; it deserves a serious referee but not acceptance as-is. read the letter →

arxiv 2504.14188 v2 pith:UA5FW5U4 submitted 2025-04-19 cs.LG

Zekai Chen , Xunkai Li , Yinlin Zhu , Rong-Hua Li , Guoren Wang This is my paper · ORCID

classification cs.LG

keywords federatedgraphlearningcondensationclient-clientcommunicationneuralnetworkspersonalizationefficiencyprivacypreservation

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proposes FedC4, a client-to-client federated graph learning method that replaces the usual exchange of raw node embeddings with graph condensation: each client compresses its private graph into a few synthetic nodes and sends those. The authors argue that this shifts federated graph collaboration from broadcast of identical messages to fine-grained, per-client tailored communication, which they categorize as a new level of personalization. They report that FedC4 beats twelve baselines on eight datasets, with an average 1.73% accuracy gain over the strongest competitor and communication savings up to 1000x. The intended payoff is a practical regime where distributed graph learning is both more accurate and cheaper to run, with privacy improved because only condensed, not raw, node information crosses the network.

What carries the argument

The load-bearing mechanism is graph condensation used as a communication format: a client's graph $G=(A,X,Y)$ is compressed into a small synthetic graph $S=(A',X',Y')$ by gradient matching, so knowledge travels between clients as synthetic node embeddings rather than raw nodes. On top of this, three modules carry the personalization: the Customizer normalizes and selectively broadcasts embedding statistics to same-cluster clients; the Node Selector groups clients by Sliced Wasserstein distance and picks representative nodes by embedding-prototype similarity; and the Graph Rebuilder reconstructs an adjacency matrix for selected nodes through sparsified self-expressive reconstruction.

What would settle it

Measure the maximum change in the transmitted condensed embeddings when one node is removed from a client graph of size $n$ while keeping the condensation ratio fixed (e.g., 2%); if that change does not decrease as $n$ grows, or if a membership-inference classifier can detect which nodes were used to build the condensed graph above chance, the paper's privacy conclusion would be refuted.

Watch

Extended reading notes

Core claim

The paper's central claim is that the bottleneck in client-client federated graph learning is message granularity, not aggregation. FedC4 replaces raw node-level exchanges with locally condensed synthetic graphs and makes every message client-specific through three modules, turning broadcast-style client-client communication into fine-grained personalized communication. The authors report that on eight datasets spanning transductive, inductive, heterogeneous, and large-scale settings, FedC4 outperforms twelve baselines while cutting communication cost by up to three orders of magnitude.

Load-bearing premise

The privacy proof assumes that the per-sample gradient bound $O(1/n)$ propagates through an inverse-Hessian factor to the synthetic embeddings, and assumes client subgraphs are disjoint; with fixed condensation ratios the resulting $O(m/n)$ bound is constant, not shrinking as graphs grow.

Editorial extensions

If this is right

Communication cost drops from $O(C^2 N d)$ for broadcast client-client exchange to $O(C \log C\, N' d)$ with $N' \ll N$, because only condensed embeddings and statistics cross the network.
Fine-grained client-client collaboration outperforms both server-client and broadcast client-client methods across transductive, inductive, heterogeneous, and large-scale settings, with an average 1.73% gain over the strongest baseline.
Condensation ratios as small as 0.1% on large graphs such as Arxiv, Products, and Reddit can match or beat baselines that otherwise run out of memory or time.
Because only synthetic embeddings are transmitted and single-node influence is argued to be bounded, privacy exposure moves from original nodes to condensed nodes, and the method tolerates added Laplace noise without collapsing accuracy.
The framework keeps accuracy stable as the number of clients grows from 5 to 15, indicating robustness to data-partition granularity.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A sharper empirical test of the privacy claim would be a membership-inference attack on the condensed embeddings at increasing original-graph sizes; because the condensation ratios are fixed (e.g., 8%, 4%, 2%), the $O(m/n)$ bound may be constant rather than diminishing, and the paper's Laplace-noise robustness curves do not directly measure membership leakage.
The Customizer/Node Selector communication template, clustering clients by distribution distance and then sending tailored statistics rather than broadcasts, could transfer to non-graph federated learning of embeddings or features.
Since the Graph Rebuilder visibly repairs condensation-induced structural damage (lower KL divergence and density, higher homophily), the same reconstruction loss could improve centralized graph condensation even without federation.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Two things to know about FedC4. The empirical work is substantial: it actually builds graph condensation into client-to-client federated graph learning, with three modules (customizer, node selector, graph rebuilder) that turn a broadcast-heavy C-C design into selective, personalized exchanges. Eight datasets, a dozen baselines, ablations, hyperparameter sweeps, and code. That is real engineering, and the accuracy and communication gains are plausible. The second thing is that the theory does not hold. The privacy proof in Sec 4.1 is a non-sequitur: Eq. (16) is a loss on the original graph, removing a node gives the standard O(1/n) gradient-difference bound, and then Eq. (21) jumps to a bound on synthetic-embedding differences through an undefined Hessian. Nothing in the proof differentiates the synthesized graph with respect to the original node removal through the condensation optimization. The 'm' factor in Eq. (22) appears from nowhere, and because the experiments use fixed condensation ratios, m/n is constant, so the claim that influence 'diminishes as the size of the original graph increases' does not describe the evaluated settings. The paper's own caveat about non-overlapping subgraphs narrows the claim further. This is load-bearing because privacy is a headline contribution. The communication complexity O(C log C) is also asserted, not derived; nothing about SWD clustering guarantees log C groups, and the server-side aggregation cost is omitted. Table 1 is messy: for some rows baselines are missing, and the 1.73% average gain is not transparently computed, though the overall win pattern looks real. Where does that leave us? The core idea—condense locally, exchange synthetic graphs selectively, rebuild graphs at the receiver—is worth having in the literature, and the empirical evaluation is better than most FGL papers. But the theoretical contributions as written are not established, and the authors oversell them. A serious referee would send this back for major revision: redo or weaken the privacy analysis (synthetic embeddings likely reduce exposure in practice, but that is a claim about attack surface, not a proven bound), derive or soften the communication complexity, and clean up the tables. The system itself is plausible and reproducible enough to engage with. So: worth a serious referee, worth reading if you work on FGL, but do not trust the proof.

Referee Report

3 major / 6 minor

Summary. The manuscript proposes FedC4, a client-to-client federated graph learning framework in which each client condenses its local graph into a small synthetic graph and then selectively shares statistics, representative node embeddings, and reconstructed graph topology with other clients through three modules: Customizer (CM), Node Selector (NS), and Graph Rebuilder (GR). The authors categorize prior FGL work into server-client and client-client paradigms, argue that fine-grained personalized client-client communication is preferable, and claim that graph condensation reduces communication cost and provides intrinsic privacy protection. They report experiments on eight graph datasets showing average accuracy improvements over baselines and large communication/training cost reductions, together with ablations and hyperparameter studies.

Significance. If the empirical claims hold, FedC4 is a potentially useful engineering contribution: it integrates graph condensation with selective client-client communication, releases code, and evaluates on a broad set of transductive, inductive, heterogeneous, and large-scale benchmarks. The proposed categorization of FGL into S-C and C-C paradigms with levels of personalization is a helpful organizing viewpoint. However, the paper's second central claim—that graph condensation is intrinsically privacy-preserving and that this is proven in Sec. 4.1—is not supported. The privacy proof is invalid as written, and the experimental privacy study in Sec. 5.5 does not test privacy. Because the privacy guarantee is advertised in the abstract, conclusion, and Q4, this is a load-bearing weakness rather than a presentation issue.

major comments (3)

[Sec. 4.1, Eqs. (16)-(22)] The privacy proof does not connect the original-graph gradient bound to the change in the synthetic graph. L(S,G) in Eq. (16) depends only on the original graph; the synthetic graph S appears only as a placeholder, so the leave-one-out gradient bound in Eqs. (18)-(20) is a standard finite-sample property of the original loss and does not bound how the condensed graph changes when an original node is removed. Eq. (21) then asserts Δ_j = ||H_syn − H_syn^{-j}|| ≤ ||H^{-1}||·||Δ∇|| with an undefined Hessian H and no influence-function argument that differentiates the condensation optimization in Eq. (6); the factor m in Eq. (22) appears without derivation. Even if such an expansion were intended, it would require uniform invertibility of H and smooth dependence of the condensed optimum on G, neither of which is stated or shown. Finally, because the experiments fix condensation ratios (8%, 4%, 2%, 0.4%, etc.), m/n is constant in the tested regime, so the conclusion that the influence of a node 'diminishes as the size of the original graph increases' does not follow from the settings actually evaluated. The caveat limiting the proof to non-overlapping subgraphs narrows the claim but does not repair the missing derivation.
[Sec. 5.5, Fig. 7(b)] The privacy evaluation does not measure privacy. Injecting Laplace noise into the condensation process and reporting that accuracy degrades slowly is a robustness-to-noise experiment, not a test of membership inference, attribute inference, or reconstruction resistance, and no formal privacy accounting links the noise scale to a guarantee. The reference to 'Privacy for Free' [2] concerns image dataset condensation and does not automatically transfer to graph embeddings, particularly because the extension proof in Sec. 4.1 is incomplete. The claim in the conclusion that GC 'mitigates privacy risks' is therefore not supported by the experimental section as written.
[Sec. 4.2, Table 2] The communication cost O(C log C N'd) for FedC4 is not derived. The manuscript describes selective broadcasting to same-cluster clients (Eq. (11), Algorithm 2), but it does not specify a clustering or communication hierarchy that would introduce a log factor; in the worst case, when all clients fall into one cluster, the cost is O(C^2 N'd). The table should either state the structural assumptions under which O(C log C N'd) holds or present the worst-case cost.

minor comments (6)

[Sec. 5.1 and Table 1] The baseline description names SGDD as one of the three FL+GC methods, but Table 1 lists a column labeled 'DosCond' and no reference for DosCond is provided; please align the text, table, and reference list.
[Eq. (12)] The Sliced Wasserstein Distance formula uses the same symbol θ both as the integration variable over the sphere and inside the supremum; as written the outer integral is redundant and the expression is not the standard SWD. Please rewrite with distinct variables or cite the precise definition used in the implementation.
[References [4] and [5]] References [4] and [5] are identical (both point to the same ICDE 2024 paper), while the text uses them for different purposes (MCond and self-expressive reconstruction); please correct the citations.
[Sec. 5.4 and Fig. 5] The text says Fig. 5(a) shows the impact of varying α and β and Fig. 5(b) shows the effect of τ, but the figure panels are labeled 'Node Selector' and 'Graph Rebuilder' and appear to show the opposite; please align the text with the figure.
[Algorithm 2] The clustering threshold δ in Algorithm 2 is not specified or studied in the hyperparameter analysis; please report its value and sensitivity.
[Abstract and Sec. 1] The claimed 'efficiency improvement up to 1000x' is not quantified in the main text; Fig. 6 shows accuracy-versus-time curves, but the basis for the 1000x factor should be made explicit.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: FedC4's empirical results are self-contained against held-out data, and the privacy proof's gap is an unsupported step rather than a circular reduction.

full rationale

I checked the derivation chain for equivalences between claimed predictions and inputs. The FedC4 pipeline uses local graph condensation by gradient matching (Eq. 6) on the local training loss; synthetic graphs are not fitted to test labels, and the reported accuracies are on held-out splits, so the performance gain is not a fitted-input-called-prediction. The privacy analysis (Sec. 4.1) is the only place where a claimed 'first-principles' result could hide circularity, but its defect is a missing derivation rather than a self-referential one: Eq. (16) defines L(S,G) without any dependence on S, Eq. (20) bounds the leave-one-out sensitivity of the original-graph loss, and Eq. (21) then asserts, without proof, that this bounds the synthetic embeddings; with FedC4's fixed condensation ratios the O(m/n) conclusion is also not the O(1/n) they state. This is an unsupported/invalid proof step, which is a correctness risk, not a circular reduction of the conclusion to its premises. Self-citations (FedGTA [10], FedTAD [35]) appear only as S-C comparison baselines in Table 1 and do not carry the derivation; there is no uniqueness theorem or ansatz imported from the authors' prior work. I therefore find no circularity step that meets the required quote-and-reduction standard.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

No new physical entities or mediators are introduced. The framework relies on standard graph condensation objectives, distance-based clustering, and prototype statistics. The main unstated costs are the unreported clustering threshold and the assumption that the condensed graph size is small enough to make the privacy bound meaningful.

free parameters (5)

NS similarity threshold tau = 0.38
Controls which nodes are selected for sharing in the Node Selector module; tuned on validation and reported in Sec 5.1.
GR loss weight alpha = 150
Balances the feature reconstruction term in Eq. 15; chosen by hyperparameter study.
GR loss weight beta = 250
Balances the sparsity penalty in Eq. 15; chosen by hyperparameter study.
SWD clustering threshold delta_cluster = not reported
Algorithm 2 assigns clients to clusters when the Sliced Wasserstein Distance is below this delta, but the value is never specified in the experiments, even though the selective communication pattern depends on it.
Condensation ratio per dataset = e.g., 8%, 4%, 2% for Cora
The number of synthetic nodes per client is a protocol choice that trades off accuracy against communication; it is not inferred from data but is a design parameter.

assumptions (4)

domain assumption Gradient matching in GCond transfers knowledge from the original graph to the condensed graph.
Eq. 6 is adopted from GCond without independent verification in the federated multi-client setting; the whole framework assumes this transfer is effective when each client condenses locally.
domain assumption Client subgraphs are non-overlapping.
Stated at the end of Sec 4.1, the privacy proof only applies when nodes do not overlap across subgraphs; real federated systems with shared entities violate this.
standard math Sliced Wasserstein Distance is a valid and reliable measure of client distribution similarity for clustering.
Used in Eq. 12 to form client clusters; the clustering quality and its effect on the selective communication strategy is not analyzed.
domain assumption The self-expressive reconstruction model underlies the GR module.
Eq. 15 assumes node features lie in a union of subspaces so that X is approximately XZ; this is imported from Gao et al. (2024), cited as reference [5].

how reviews work

0 comments

Cite this review

Pith. "Pith review of Rethinking Client-oriented Federated Graph Learning." pith.science (2026). https://pith.science/paper/UA5FW5U4

@misc{pith2026250414188,
  author       = {Pith},
  title        = {Pith review of: Rethinking Client-oriented Federated Graph Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UA5FW5U4}},
  note         = {Machine review of arXiv:2504.14188}
}

read the original abstract

As a new distributed graph learning paradigm, Federated Graph Learning (FGL) facilitates collaborative model training across local systems while preserving data privacy. We review existing FGL approaches and categorize their optimization mechanisms into: (1) Server-Client (S-C), where clients upload local model parameters for server-side aggregation and global updates; (2) Client-Client (C-C), which allows direct exchange of information between clients and customizing their local training process. We reveal that C-C shows superior potential due to its refined communication structure. However, existing C-C methods broadcast redundant node representations, incurring high communication costs and privacy risks at the node level. To this end, we propose FedC4, which combines graph Condensation with C-C Collaboration optimization. Specifically, FedC4 employs graph condensation technique to refine the knowledge of each client's graph into a few synthetic embeddings instead of transmitting node-level knowledge. Moreover, FedC4 introduces three novel modules that allow the source client to send distinct node representations tailored to the target client's graph properties. Experiments on eight public real-world datasets show that FedC4 outperforms state-of-the-art baselines in both task performance and communication cost. Our code is now available on https://github.com/Ereshkigal1/FedC4.

Figures

Figures reproduced from arXiv: 2504.14188 by the authors.

**Figure 2.** Overview of our proposed FedC4 framework. The left side illustrates the overall workflow of server-side broadcasting [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 4.** (a) Ablation Study of FedC4 on CM Module. (b) [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗

Figures from the paper (2 more)

**Figure 5.** Figure 5: Hyper-parameter study for (a) 𝜏 in Node Selector Module. (b) 𝛼 , 𝛽 in Graph Rebuilder Module. (a) Ogbn-arxiv (b) Ogbn-products [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

**Figure 6.** Figure 6: Efficiency evaluation on Arxiv and Products. [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 8 canonical work pages

[2]

Tian Dong, Bo Zhao, and Lingjuan Lyu. 2022. Privacy for free: How does dataset condensation help privacy?. In International Conference on Machine Learning . PMLR, 5378–5396

2022
[1]

Joan Bruna, Wojciech Zaremba, Arthur Szlam, and Yann LeCun. 2013. Spectral net- works and locally connected networks on graphs. arXiv preprint arXiv:1312.6203 (2013)

arXiv 2013
[3]

Liang Gao, Huazhu Fu, Li Li, Yingwen Chen, Ming Xu, and Cheng-Zhong Xu
[5]

Xinyi Gao, Tong Chen, Yilong Zang, Wentao Zhang, Quoc Viet Hung Nguyen, Kai Zheng, and Hongzhi Yin. 2024. Graph condensation for inductive node repre- sentation learning. In 2024 IEEE 40th International Conference on Data Engineering (ICDE). IEEE, 3056–3069

2024
[6]

Will Hamilton, Zhitao Ying, and Jure Leskovec. 2017. Inductive representation learning on large graphs. Advances in neural information processing systems 30 (2017)

2017
[7]

Wenke Huang, Guancheng Wan, Mang Ye, and Bo Du. 2024. Federated graph semantic and structural learning. arXiv preprint arXiv:2406.18937 (2024)

arXiv 2024
[8]

Wei Jin, Lingxiao Zhao, Shichang Zhang, Yozen Liu, Jiliang Tang, and Neil Shah. 2021. Graph condensation for graph neural networks. arXiv preprint arXiv:2110.07580 (2021)

arXiv 2021
[9]

Kipf and Max Welling

Thomas N. Kipf and Max Welling. 2017. Semi-Supervised Classification with Graph Convolutional Networks. arXiv:1609.02907 [cs.LG] https://arxiv.org/abs/ 1609.02907

arXiv 2017

Show all 34 references

[10]

Xunkai Li, Zhengyu Wu, Wentao Zhang, Yinlin Zhu, Rong-Hua Li, and Guoren Wang. 2024. Fedgta: Topology-aware averaging for federated graph learning. arXiv preprint arXiv:2401.11755 (2024)

2024 arXiv
[11]

Andreas Loukas. 2019. Graph reduction with spectral and cut guarantees.Journal of Machine Learning Research 20, 116 (2019), 1–42

2019
[12]

Brendan McMahan, Eider Moore, Daniel Ramage, Seth Hampson, and Blaise Aguera y Arcas. 2017. Communication-efficient learning of deep net- works from decentralized data. In Artificial intelligence and statistics . PMLR, 1273–1282

2017
[13]

Oleg Platonov, Denis Kuznedelev, Michael Diskin, Artem Babenko, and Liudmila Prokhorenkova. 2023. A critical look at the evaluation of GNNs under heterophily: Are we really making progress? arXiv preprint arXiv:2302.11640 (2023)

2023 arXiv
[14]

Yuxin Qiu. 2023. Default Risk Assessment of Internet Financial Enterprises Based on Graph Neural Network. In 2023 IEEE 6th Information Technology, Networking, Electronic and Automation Control Conference , Vol. 6. IEEE, 592–596

2023
[15]

Zongshuai Qu, Tao Yao, Xinghui Liu, and Gang Wang. 2023. A Graph Con- volutional Network Based on Univariate Neurodegeneration Biomarker for Alzheimer’s Disease Diagnosis. IEEE Journal of Translational Engineering in Health and Medicine (2023)

2023
[16]

Maria Rigaki and Sebastian Garcia. 2023. A survey of privacy attacks in machine learning. Comput. Surveys 56, 4 (2023), 1–34

2023
[17]

Herbert Robbins and Sutton Monro. 1951. A stochastic approximation method. The annals of mathematical statistics (1951), 400–407

1951
[18]

Oleksandr Shchur, Maximilian Mumme, Aleksandar Bojchevski, and Stephan Günnemann. 2018. Pitfalls of graph neural network evaluation. arXiv preprint arXiv:1811.05868 (2018)

2018 arXiv
[19]

Petar Velickovic, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Lio, Yoshua Bengio, et al. 2017. Graph attention networks. stat 1050, 20 (2017), 10–48550

2017
[20]

Max Welling. 2009. Herding dynamical weights to learn. In Proceedings of the 26th annual international conference on machine learning . 1121–1128

2009
[21]

Shiwen Wu, Fei Sun, Wentao Zhang, Xu Xie, and Bin Cui. 2022. Graph neural networks in recommender systems: a survey. Comput. Surveys 55, 5 (2022), 1–37

2022
[22]

Zonghan Wu, Shirui Pan, Fengwen Chen, Guodong Long, Chengqi Zhang, and S Yu Philip. 2020. A comprehensive survey on graph neural networks. IEEE transactions on neural networks and learning systems 32, 1 (2020), 4–24

2020
[23]

Han Xie, Jing Ma, Li Xiong, and Carl Yang. 2021. Federated graph classification over non-iid graphs. Advances in neural information processing systems 34 (2021), 18839–18852

2021
[24]

Han Xie, Li Xiong, and Carl Yang. 2023. Federated node classification over graphs with latent link-type heterogeneity. In Proceedings of the ACM Web Conference

2023
[25]

Bo Yan. 2024. Federated Graph Condensation with Information Bottleneck Principles. arXiv preprint arXiv:2405.03911 (2024)

2024 arXiv
[26]

Beining Yang, Kai Wang, Qingyun Sun, Cheng Ji, Xingcheng Fu, Hao Tang, Yang You, and Jianxin Li. 2024. Does graph distillation see like vision dataset counterpart? Advances in Neural Information Processing Systems 36 (2024)

2024
[28]

Zhilin Yang, William Cohen, and Ruslan Salakhudinov. 2016. Revisiting semi- supervised learning with graph embeddings. In International conference on ma- chine learning. PMLR, 40–48

2016
[29]

Yuhang Yao, Weizhao Jin, Srivatsan Ravi, and Carlee Joe-Wong. 2024. FedGCN: Convergence-communication tradeoffs in federated training of graph convolu- tional networks. Advances in neural information processing systems 36 (2024)

2024
[30]

Hanqing Zeng, Hongkuan Zhou, Ajitesh Srivastava, Rajgopal Kannan, and Viktor Prasanna. 2019. Graphsaint: Graph sampling based inductive learning method. arXiv preprint arXiv:1907.04931 (2019)

2019 arXiv
[31]

Ke Zhang, Lichao Sun, Bolin Ding, Siu Ming Yiu, and Carl Yang. 2024. Deep Efficient Private Neighbor Generation for Subgraph Federated Learning. In Pro- ceedings of the 2024 SIAM International Conference on Data Mining (SDM) . SIAM, 806–814

2024
[32]

Ke Zhang, Carl Yang, Xiaoxiao Li, Lichao Sun, and Siu Ming Yiu. 2021. Sub- graph federated learning with missing neighbor generation. Advances in Neural Information Processing Systems 34 (2021), 6671–6682

2021
[33]

Xin Zheng, Miao Zhang, Chunyang Chen, Quoc Viet Hung Nguyen, Xingquan Zhu, and Shirui Pan. 2024. Structure-free graph condensation: From large-scale graphs to condensed graph-free data. Advances in Neural Information Processing Systems 36 (2024)

2024
[34]

Jie Zhou, Ganqu Cui, Shengding Hu, Zhengyan Zhang, Cheng Yang, Zhiyuan Liu, Lifeng Wang, Changcheng Li, and Maosong Sun. 2020. Graph neural networks: A review of methods and applications. AI open 1 (2020), 57–81

2020
[35]

Yinlin Zhu, Xunkai Li, Zhengyu Wu, Di Wu, Miao Hu, and Rong-Hua Li. 2024. FedTAD: Topology-aware Data-free Knowledge Distillation for Subgraph Feder- ated Learning. arXiv preprint arXiv:2404.14061 (2024)

2024 arXiv
[2022]

In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Feddc: Federated learning with non-iid data via local drift decoupling and correction. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 10112–10121

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.

[1] [2]

Tian Dong, Bo Zhao, and Lingjuan Lyu. 2022. Privacy for free: How does dataset condensation help privacy?. In International Conference on Machine Learning . PMLR, 5378–5396

2022

[2] [1]

Joan Bruna, Wojciech Zaremba, Arthur Szlam, and Yann LeCun. 2013. Spectral net- works and locally connected networks on graphs. arXiv preprint arXiv:1312.6203 (2013)

arXiv 2013

[3] [3]

Liang Gao, Huazhu Fu, Li Li, Yingwen Chen, Ming Xu, and Cheng-Zhong Xu

[4] [5]

Xinyi Gao, Tong Chen, Yilong Zang, Wentao Zhang, Quoc Viet Hung Nguyen, Kai Zheng, and Hongzhi Yin. 2024. Graph condensation for inductive node repre- sentation learning. In 2024 IEEE 40th International Conference on Data Engineering (ICDE). IEEE, 3056–3069

2024

[5] [6]

Will Hamilton, Zhitao Ying, and Jure Leskovec. 2017. Inductive representation learning on large graphs. Advances in neural information processing systems 30 (2017)

2017

[6] [7]

Wenke Huang, Guancheng Wan, Mang Ye, and Bo Du. 2024. Federated graph semantic and structural learning. arXiv preprint arXiv:2406.18937 (2024)

arXiv 2024

[7] [8]

Wei Jin, Lingxiao Zhao, Shichang Zhang, Yozen Liu, Jiliang Tang, and Neil Shah. 2021. Graph condensation for graph neural networks. arXiv preprint arXiv:2110.07580 (2021)

arXiv 2021

[8] [9]

Kipf and Max Welling

Thomas N. Kipf and Max Welling. 2017. Semi-Supervised Classification with Graph Convolutional Networks. arXiv:1609.02907 [cs.LG] https://arxiv.org/abs/ 1609.02907

arXiv 2017