Pith. sign in

REVIEW 4 cited by

Graph Condensation for Graph Neural Networks

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2110.07580 v4 pith:EXKGQBUA submitted 2021-10-14 cs.LG cs.AI

Wei Jin , Lingxiao Zhao , Shichang Zhang , Yozen Liu , Jiliang Tang , Neil Shah This is my paper · ORCID
classification cs.LGcs.AI
keywords graphcondensationgraphsneuraloriginalconcernscondensegnns
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Given the prevalence of large-scale graphs in real-world applications, the storage and time for training neural models have raised increasing concerns. To alleviate the concerns, we propose and study the problem of graph condensation for graph neural networks (GNNs). Specifically, we aim to condense the large, original graph into a small, synthetic and highly-informative graph, such that GNNs trained on the small graph and large graph have comparable performance. We approach the condensation problem by imitating the GNN training trajectory on the original graph through the optimization of a gradient matching loss and design a strategy to condense node futures and structural information simultaneously. Extensive experiments have demonstrated the effectiveness of the proposed framework in condensing different graph datasets into informative smaller graphs. In particular, we are able to approximate the original test accuracy by 95.3% on Reddit, 99.8% on Flickr and 99.0% on Citeseer, while reducing their graph size by more than 99.9%, and the condensed graphs can be used to train various GNN architectures.Code is released at https://github.com/ChandlerBang/GCond.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 4 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Scalable Topology-Preserving Graph Coarsening: Concepts and Algorithms

    cs.LG 2026-01 conditional novelty 6.0 of 10

    STPGC coarsens graphs by iteratively removing dominated nodes and edges (plus a neighborhood-coning step), preserving clique-complex homotopy and yielding 8.7–37x speedups over GEC on large graphs.

  2. Dynamic Graph Condensation

    cs.LG 2025-06 conditional novelty 6.0 of 10

    DyGC, the first framework for dynamic graph condensation, synthesizes a small temporal graph that preserves enough spatiotemporal structure to train dynamic GNNs with up to 1846 times speedup and around 96 percent fidelity.

  3. Simple yet Effective Graph Distillation via Clustering

    cs.LG 2025-05 conditional novelty 6.0 of 10

    ClustGDD distills large graphs by clustering node embeddings and refining synthetic attributes, achieving state-of-the-art node classification accuracy at orders of magnitude lower time cost.

  4. GCAL: Adapting Graph Models to Evolving Domain Shifts

    cs.LG 2025-05 conditional novelty 5.0 of 10

    GCAL combines information-maximization adaptation with variational memory graph generation to prevent catastrophic forgetting in unsupervised continual graph domain adaptation.

Pith tools