Pith. sign in

REVIEW 3 major objections 5 minor 78 references

Graph Structure Refinement with Energy-based Contrastive Learning

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read ECL-GSR claims state-of-the-art node classification on eight benchmarks by repairing graph structure with an energy-based contrastive objective, and reports faster training with fewer samples and less memory.

desk verdict Strong empirical structure-refinement results undercut by a load-bearing mismatch: the implemented generative loss never uses the SGLD samples the theory requires. read the letter →

arxiv 2412.17856 v3 pith:UCUR6W2H submitted 2024-12-20 cs.LG

Xianlin Zeng , Yufeng Wang , Yuqi Sun , Guodong Guo , Wenrui Ding , Baochang Zhang This is my paper · ORCID
classification cs.LG
keywords graphstructurerefinementenergy-basedmodelscontrastivelearningneuralnetworksnodeclassificationunsupervisedrepresentationstochasticgradientLangevindynamicsedgeprediction
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

Real-world graphs arrive with missing and spurious edges, and message-passing GNNs amplify that noise. This paper proposes ECL-GSR, an unsupervised structure-refinement module that learns node representations with an energy-based contrastive objective and then rewires the graph by adding or deleting edges according to representation similarity. The paper claims that on eight node-classification benchmarks this refined graph yields state-of-the-art accuracy, with faster convergence and lower memory than leading structure-learning baselines, and that the method is especially strong with very few training labels and under random edge perturbations. It also argues that the standard contrastive loss is a special case of the proposed energy-based objective when the generative term is switched off. The combination matters because structure refinement is a bottleneck for deploying GNNs on noisy or incompletely observed networks.

What carries the argument

The engine is the ECL loss built on the energy $E_\theta(\nu,\nu') = \lVert z - z' \rVert^2 / \tau$ over pairs of augmented views of a dual-attribute graph, whose node features concatenate raw attributes with DeepWalk structural embeddings. Bayes' rule decomposes the objective into the discriminative conditional likelihood and the generative marginal likelihood; the former is implemented as the contrastive loss of Eq. (15), the latter as the positive-pair log-sum-exp of Eq. (16), and SGLD is the sampler behind Eq. (13). A normalized-cosine edge predictor plus relaxed Bernoulli sampling turns learned representation similarities into a refined adjacency matrix, and the full objective adds an L2 energy regularizer and a cross-entropy classification term.

What would settle it

Train the same ECL-GSR pipeline twice on Cora, once with the implemented generative loss of Eq. (16) and once with the SGLD-based gradient of Eq. (13) using the sampled views $\nu^*$ from Algorithm 1; if the two versions reach nearly identical accuracy and refined graphs, then the implemented objective does not actually depend on the energy-model sampler, and the paper's generative-training claim is not doing the work attributed to it.

Watch

Extended reading notes

Core claim

ECL-GSR models the joint distribution of two augmented views of a graph as an energy-based model, with energy given by the squared distance between their representations divided by a temperature. Using Bayes' rule, the negative log-likelihood separates into a conditional term, which becomes a SimCLR-style contrastive loss, and a marginal term, which acts as a generative loss shaped by stochastic-gradient Langevin dynamics; the paper proves that the marginal distribution of a view is itself an energy-based model. After this representation learning, an edge predictor computes normalized cosine similarities between node embeddings and binarizes them with relaxed Bernoulli sampling to produce a refined adjacency matrix, and a three-layer GNN classifies nodes on the refined graph. The paper's empirical claim is that this two-step pipeline outperforms thirteen structure-learning baselines on eight datasets, and its theoretical claim is that the discriminative contrastive loss is a special case of the ECL objective when the generative term is removed.

Load-bearing premise

The load-bearing premise is that the approximated generative term in Eq. (16) delivers the same benefit as actually sampling from the energy model in Eq. (13), even though the implemented loss never uses the sampled views that Algorithm 1 produces.

Editorial extensions

If this is right

  • If the central claim holds, noisy or incomplete graphs can be repaired without labels, and the repaired graph alone is enough to lift GNN node-classification accuracy above previous structure-learning methods on eight benchmarks.
  • The reported label-efficiency results mean that in semi-supervised settings with 1% training data, ECL-GSR outperforms baselines, so the method is a candidate for graphs with very few labels.
  • Because training uses mini-batch subgraphs and only three SGLD steps, the approach scales to large graphs like Pubmed and OGB-Arxiv, where several competing methods run out of memory.
  • The ablation studies imply that the value is in the combination: disabling either the generative or the discriminative term measurably hurts accuracy.
  • The robustness experiments suggest that the refined structure survives random edge additions and removals up to 80% intensity, and the learned graph keeps inter-class edges sparser than intra-class edges.

Reading between the lines

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

  • Inference beyond the paper: because Eq. (16), the implemented generative loss, contains no dependence on the views $\nu^*$ produced by Algorithm 1's SGLD loop, the practical method may be equivalent to a SimCLR-style contrastive loss plus an L2 regularizer on positive-pair energies; an ablation that replaces Eq. (16) with the true SGLD gradient would settle whether the energy-based sampler contrib
  • Inference beyond the paper: the similarity-based edge-prediction rule is task-agnostic, so the framework should transfer to link prediction and graph classification; the paper's appendix results on three graph-classification datasets are consistent with, but do not prove, that transfer.
  • Inference beyond the paper: on the three heterophilic webpage graphs, where raw homophily is below 0.2, the method still gains, suggesting the dual-attribute features learn semantic similarity that overrides noisy topology; a direct test would compare the method's added edges against oracle within-class edges to identify whether gains come from adding intra-class links or deleting inter-class nois
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes ECL-GSR, a graph structure refinement framework that combines energy-based models (EBMs) with contrastive learning. The method builds a joint distribution over augmented graph views using an energy function, decomposes the objective into discriminative and generative terms, and uses the learned representations to add or remove edges before training a GNN classifier. The authors report state-of-the-art node classification accuracy on eight benchmark datasets, faster training and lower memory use than leading baselines, plus ablations, robustness tests, statistical tests, and a graph-classification extension. The central theoretical claim is that the generative EBM term is trained via SGLD, making this the first EBM-plus-contrastive-learning method for graph structure refinement.

Significance. If the theoretical claim held, the paper would offer a simple and efficient structure-refinement module with broad applicability: a single unsupervised-style loss that improves GNN node classification across small and large graphs. The empirical package is a genuine strength: eight datasets, standard splits and train-ratio experiments, comparisons against 13 baselines, efficiency and scalability analysis, robustness under edge perturbations, component ablations, and statistical significance tests. The low-resource claims are also interesting and potentially useful. However, the core novelty rests on the generative EBM objective, and the implemented loss in Eq. (16) is not the SGLD-based objective derived in Eq. (13). The paper also does not release code and leaves the graph augmentation operator underspecified, which limits verifiability. The significance is therefore conditional: the empirical results may stand, but the stated theoretical contribution needs either repair or a substantially reframed presentation.

major comments (3)
  1. [Methodology, Eqs. (13)-(16) and Algorithm 1] The implemented generative loss is disconnected from the SGLD sampling required by the theory. Algorithm 1 samples {ν*_n} from p_d(ν), runs K SGLD updates via Eq. (3), and then directs the reader to compute the generative term with Eq. (16). However, Eq. (16) is a log-sum-exp over exp(-||z_n - z'_n||^2/τ), where z_n = f_θ(ν_n) and z'_n = f_θ(ν'_n) are the original augmented views; the refined samples ν* never appear in the loss. Consequently, the negative phase E_{p_θ}[∇_θ E_θ(ν)] in Eq. (13) is absent from the optimized objective, and Eq. (13) is not the gradient of what is implemented. The implemented term is instead a softmin over the N positive-pair energies. Please either modify the algorithm so that ν* is actually used in the generative term, provide a proof that Eq. (16) is a valid surrogate for the SGLD-based objective, or explicitly reframe the contribution as contrastive graph structure refinement with a positive-pair log-sum-exp auxiliary loss.
  2. [Abstract and Training Objective, Eq. (19) and Algorithm 1] The claim that ECL-GSR is an unsupervised method is contradicted by the training protocol. Eq. (19) minimizes L_E(θ) + µ L_C(θ), where L_C is the cross-entropy loss computed with the labeled nodes Y, and Algorithm 1 explicitly takes Y as input and updates both θ_E and θ_C. Since the encoder f_θ producing the refined structure is updated with L_C, the structure refinement in the reported experiments is supervised. The 'unsupervised' claim should be restricted to the ECL component, or the authors should report a variant trained without L_C to substantiate the abstract's wording.
  3. [Preprocessing and Implementation Details, §3.1 and §4.1] The data augmentation operator T is described only as 'a random Gaussian blur' in Section 4.1. For graph data it is not specified whether the blur applies to node features, to the adjacency matrix, or to both, nor how the two views t(g†) and t'(g†) differ. This makes the construction of positive and negative pairs, and the domain on which SGLD in Eq. (3) runs, underspecified. The ambiguity is compounded by the absence of released code. Please provide the exact augmentation recipe, including any parameters and whether it is topology-aware.
minor comments (5)
  1. [Eq. (16) and surrounding text] In the simplification preceding Eq. (16), the expression p_hat_theta(ν_n) = (1/N) Σ_{n=1}^N p_theta(ν_n, ν'_n) reuses the index n on both sides; a different summation index would make the averaging clear.
  2. [Eq. (14)-(15)] The text says there are N positive and 2(N-1) negative samples, but the denominator of Eq. (15) appears to include the positive pair among the N views ν'_m; please clarify the counting so the denominator matches the SimCLR-style softmax.
  3. [Figure 2] The efficiency and scalability figure is described as 'nearer to the upper left corner signifies superior overall performance,' but the axes and plotted quantities are not labeled; please define the axes and units.
  4. [Table 3] The pairwise t-test description does not state whether the compared results are paired or independent across runs, and several P-values are extremely small despite modest accuracy differences; please report effect sizes or confidence intervals.
  5. [Appendix, Training Stability] In the training-stability subsection, the text states α and β are set to 0.1 and 0.001, whereas the main text reports α=0.1, β=0.01; please reconcile these values.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the ECL-GSR derivation is self-contained and benchmarked externally; the Eq. (16)/SGLD gap is a correctness concern, not a circular reduction.

full rationale

The paper's derivation chain is not circular in the sense defined by the review criteria. The EBM construction (Eqs. 5-13) is a standard marginalization-plus-Bayes derivation: a joint energy-based distribution is defined, the marginal is shown to be an EBM, and the objective is decomposed into a conditional (discriminative) term and a marginal (generative) term. No fitted parameter is renamed as a prediction, and no benchmark accuracy is used as an input to the loss. The statement that the discriminative loss is a special case of the ECL loss when α=0 is a definitional property of the composite loss, but the paper does not use that property as evidence of empirical success; it is presented as a structural observation. The empirical claims are evaluated against external baselines on standard datasets, with uniform hyperparameters, so the results are not reverse-engineered from the target quantities. The main weakness is that Algorithm 1 samples ν* via SGLD, but Eq. (16) never uses ν*; the implemented 'generative' term is instead a positive-pair log-sum-exp. This is a serious mismatch between the stated generative-training theory and the actual loss, and it undermines the theoretical contribution. However, it is not circularity: Eq. (16) is not obtained by substituting the target result into the inputs, and no self-citation chain forces the conclusion. No load-bearing self-citation was identified. The cited motivation from Kim and Ye (2022) and Wang et al. (2022b) is background support, and the central equations are derived in the paper itself rather than imported as an unexamined uniqueness claim. Overall, the paper's derivation does not reduce to its own inputs by construction, so the circularity score is 0.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

The framework relies on a small set of hyperparameters tuned on the benchmarks and on several domain assumptions, the most fragile being the use of SGLD for the generative term and the cosine edge scorer. No new physical or ontological entities are introduced; ECL is a loss, not a new entity.

free parameters (7)
  • alpha (generative term weight) = 0.1
    Trade-off between discriminative and generative terms in Eq. (12); chosen by performance on benchmarks.
  • beta (energy regularizer weight) = 0.01
    Weight for L2 energy regularization in Eq. (17).
  • mu (classification loss weight) = 0.01
    Weight for supervised node classification loss in Eq. (19).
  • tau (temperature) = 0.1
    Temperature in energy function Eθ = ||z-z'||^2 / tau; controls sharpness of the distribution.
  • SGLD iterations K = 3
    Number of Langevin sampling steps; the paper says further iterations give marginal gains (Appendix).
  • hidden dimension eF = 128
    Output dimension of encoder phi_theta and projection phi_theta; ablations show lower dimensions hurt accuracy.
  • batch size N = 64
    Subgraph sample size for edge sampling; fixed for efficiency.
assumptions (6)
  • standard math Boltzmann distribution with finite partition function Z(theta) defines p_theta.
    Used in Eqs. (5)-(7) to define the joint and marginal distributions.
  • domain assumption Semantically similar views have close projections and high p_d; dissimilar views have distant projections.
    Core modeling assumption behind Definition 1 and the contrastive objective.
  • domain assumption SGLD with K=3 steps approximates sampling from p_theta.
    Convergence is only guaranteed as k goes to infinity and lambda goes to 0; the paper uses K=3.
  • ad hoc to paper Random Gaussian blur is a valid graph augmentation to generate views nu and nu'.
    No precise definition is given for Gaussian blur on graph subgraphs; the augmentation is used without ablation.
  • domain assumption Cosine similarity of ECL embeddings predicts edge existence.
    Eq. (18) uses cosine similarity to add or remove edges; the paper does not validate this choice against other edge predictors.
  • domain assumption DeepWalk structural embeddings add useful information for downstream classification.
    Preprocessing concatenates DeepWalk embeddings X_s with raw features to build the dual-attribute graph.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Graph Structure Refinement with Energy-based Contrastive Learning." pith.science (2026). https://pith.science/paper/UCUR6W2H

@misc{pith2026241217856,
  author       = {Pith},
  title        = {Pith review of: Graph Structure Refinement with Energy-based Contrastive Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UCUR6W2H}},
  note         = {Machine review of arXiv:2412.17856}
}
read the original abstract

Graph Neural Networks (GNNs) have recently gained widespread attention as a successful tool for analyzing graph-structured data. However, imperfect graph structure with noisy links lacks enough robustness and may damage graph representations, therefore limiting the GNNs' performance in practical tasks. Moreover, existing generative architectures fail to fit discriminative graph-related tasks. To tackle these issues, we introduce an unsupervised method based on a joint of generative training and discriminative training to learn graph structure and representation, aiming to improve the discriminative performance of generative models. We propose an Energy-based Contrastive Learning (ECL) guided Graph Structure Refinement (GSR) framework, denoted as ECL-GSR. To our knowledge, this is the first work to combine energy-based models with contrastive learning for GSR. Specifically, we leverage ECL to approximate the joint distribution of sample pairs, which increases the similarity between representations of positive pairs while reducing the similarity between negative ones. Refined structure is produced by augmenting and removing edges according to the similarity metrics among node representations. Extensive experiments demonstrate that ECL-GSR outperforms the state-of-the-art on eight benchmark datasets in node classification. ECL-GSR achieves faster training with fewer samples and memories against the leading baseline, highlighting its simplicity and efficiency in downstream tasks.

Figures

Figures reproduced from arXiv: 2412.17856 by the authors.

Figure 1
Figure 1. Illustration of the procedure of ECL-GSR. Preprocessed dual-attribute graph undergoes data augmentations, energy [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Training time and space analysis on Cora, Citeseer, Actor, and Pubmed datasets. [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Performance study of ECL-GSR variants and other [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (4 more)
Figure 5
Figure 5. Figure 5: Hyperparameter α and dimensionality Fe analysis of ECL-GSR on four datasets. Ablation Study (RQ3) Component analysis As illustrated in [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 7
Figure 7. Figure 7: Visualization of the refined adjacency matrixes by [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: It is observed that a few steps of SGLD iterations [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 9
Figure 9. Figure 9: Training dynamics of ECL-GSR with multiple [PITH_FULL_IMAGE:figures/full_fig_p014_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

78 extracted references · 65 canonical work pages

  1. [1]

    Bayes, T. 1763. LII. An essay towards solving a problem in the doctrine of chances. By the late Rev. Mr. Bayes, FRS communicated by Mr. Price, in a letter to John Canton, AMFR S. Philosophical Transactions of the Royal Society of London, (53): 370--418

    work page

  2. [2]

    Bo, D.; Fang, Y.; Liu, Y.; and Shi, C. 2024. Graph Contrastive Learning with Stable and Scalable Spectral Encoding. Advances in Neural Information Processing Systems, 36

    work page 2024

  3. [3]

    M.; Ong, C

    Borgwardt, K. M.; Ong, C. S.; Sch \"o nauer, S.; Vishwanathan, S.; Smola, A. J.; and Kriegel, H.-P. 2005. Protein Function Prediction via Graph Kernels. Bioinformatics, 21(suppl\_1): i47--i56

    work page 2005

  4. [4]

    Chen, T.; Bian, S.; and Sun, Y. 2019. Are powerful graph neural nets necessary? a dissection on graph classification. arXiv preprint arXiv:1905.04579

    arXiv 2019

  5. [5]

    Chen, T.; Kornblith, S.; Norouzi, M.; and Hinton, G. 2020. A Simple Framework for Contrastive Learning of Visual Representations. In International Conference on Machine Learning. PMLR

    work page 2020

  6. [6]

    Chen, Y.; Wu, L.; and Zaki, M. 2020. Iterative Deep Graph Learning for Graph Neural Networks: Better and Robust Node Embeddings. Advances in Neural Information Processing Systems, 33: 19314--19326

    work page 2020

  7. [7]

    Dai, H.; Li, H.; Tian, T.; Huang, X.; Wang, L.; Zhu, J.; and Song, L. 2018. Adversarial Attack on Graph Structured Data. In International Conference on Machine Learning. PMLR

    work page 2018

  8. [8]

    K.; Lopez de Compadre, R

    Debnath, A. K.; Lopez de Compadre, R. L.; Debnath, G.; Shusterman, A. J.; and Hansch, C. 1991. Structure-Activity Relationship of Mutagenic Aromatic and Heteroaromatic Nitro Compounds. Correlation with Molecular Orbital Energies and Hydrophobicity. Journal of Medicinal Chemistry, 34(2): 786--797

    work page 1991

Show all 78 references
  1. [9]

    Ding, K.; Li, J.; Bhanushali, R.; and Liu, H. 2019. Deep Anomaly Detection on Attributed Networks. In Proceedings of the 2019 SIAM International Conference on Data Mining. SIAM

    2019

  2. [10]

    V.; and Swami, A

    Dong, Y.; Chawla, N. V.; and Swami, A. 2017. metapath2vec: Scalable Representation Learning for Heterogeneous Networks. In Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 135--144

    2017

  3. [11]

    Du, Y.; and Mordatch, I. 2019. Implicit Generation and Modeling with Energy-Based Models. Advances in Neural Information Processing Systems, 32

    2019

  4. [12]

    K.; Maclaurin, D.; Iparraguirre, J.; Bombarell, R.; Hirzel, T.; Aspuru-Guzik, A.; and Adams, R

    Duvenaud, D. K.; Maclaurin, D.; Iparraguirre, J.; Bombarell, R.; Hirzel, T.; Aspuru-Guzik, A.; and Adams, R. P. 2015. Convolutional Networks on Graphs for Learning Molecular Fingerprints. Advances in Neural Information Processing Systems, 28

    2015

  5. [13]

    Franceschi, L.; Niepert, M.; Pontil, M.; and He, X. 2019. Learning Discrete Structures for Graph Neural Networks. In International Conference on Machine Learning. PMLR

    2019

  6. [14]

    S.; Riley, P

    Gilmer, J.; Schoenholz, S. S.; Riley, P. F.; Vinyals, O.; and Dahl, G. E. 2017. Neural Message Passing for Quantum Chemistry. In International Conference on Machine Learning. PMLR

    2017

  7. [15]

    Grathwohl, W.; Wang, K.-C.; Jacobsen, J.-H.; Duvenaud, D.; Norouzi, M.; and Swersky, K. 2020. Your Classifier is Secretly an Energy Based Model and You Should Treat it Like One. In International Conference on Learning Representations

    2020

  8. [16]

    Grover, A.; and Leskovec, J. 2016. node2vec: Scalable Feature Learning for Networks. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 855--864

    2016

  9. [17]

    Gunel, B.; Du, J.; Conneau, A.; and Stoyanov, V. 2020. Supervised Contrastive Learning for Pre-trained Language Model Fine-tuning. arXiv preprint arXiv:2011.01403

    2020 arXiv

  10. [18]

    Hassani, K.; and Khasahmadi, A. H. 2020. Contrastive Multi-view Representation Learning on Graphs. In International Conference on Machine Learning, 4116--4126. PMLR

    2020

  11. [19]

    He, K.; Fan, H.; Wu, Y.; Xie, S.; and Girshick, R. 2020. Momentum Contrast for Unsupervised Visual Representation Learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    2020

  12. [20]

    Hinton, G. E. 2002. Training Products of Experts by Minimizing Contrastive Divergence. Neural Computation, 14(8): 1771--1800

    2002

  13. [21]

    Hu, W.; Fey, M.; Zitnik, M.; Dong, Y.; Ren, H.; Liu, B.; Catasta, M.; and Leskovec, J. 2020. Open Graph Benchmark: Datasets for Machine Learning on Graphs. Advances in Neural Information Processing Systems, 33: 22118--22133

    2020

  14. [22]

    Jin, W.; Ma, Y.; Liu, X.; Tang, X.; Wang, S.; and Tang, J. 2020. Graph Structure Learning for Robust Graph Neural Networks. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

    2020

  15. [23]

    Kim, B.; and Ye, J. C. 2022. Energy-Based Contrastive Learning of Visual Representations. Advances in Neural Information Processing Systems, 35: 4358--4369

    2022

  16. [24]

    N.; and Welling, M

    Kipf, T. N.; and Welling, M. 2016. Semi-supervised Classification with Graph Convolutional Networks. arXiv preprint arXiv:1609.02907

    2016 arXiv

  17. [25]

    LeCun, Y.; Chopra, S.; Hadsell, R.; Ranzato, M.; and Huang, F. 2006. A Tutorial on Energy-based Learning. Predicting Structured Data, 1(0)

    2006

  18. [26]

    Lee, J.; Lee, I.; and Kang, J. 2019. Self-attention Graph Pooling. In International Conference on Machine Learning. PMLR

    2019

  19. [27]

    Li, K.; Liu, Y.; Ao, X.; Chi, J.; Feng, J.; Yang, H.; and He, Q. 2022. Reliable Representations Make a Stronger Defender: Unsupervised Structure Refinement for Robust GNN. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining

    2022

  20. [28]

    Li, Z.; Wang, L.; Sun, X.; Luo, Y.; Zhu, Y.; Chen, D.; Luo, Y.; Zhou, X.; Liu, Q.; Wu, S.; et al. 2023. GSLB: The Graph Structure Learning Benchmark. arXiv preprint arXiv:2310.05174

    2023

  21. [29]

    Liu, N.; Wang, X.; Wu, L.; Chen, Y.; Guo, X.; and Shi, C. 2022 a . Compact Graph Structure Learning via Mutual Information Compression. In Proceedings of the ACM Web Conference 2022

    2022

  22. [30]

    Liu, Y.; Zheng, Y.; Zhang, D.; Chen, H.; Peng, H.; and Pan, S. 2022 b . Towards Unsupervised Deep Graph Structure Learning. In Proceedings of the ACM Web Conference 2022

    2022

  23. [31]

    Namata, G.; London, B.; Getoor, L.; Huang, B.; and Edu, U. 2012. Query-driven Active Surveying for Collective Classification. In 10th International Workshop on Mining and Learning with Graphs, volume 8

    2012

  24. [32]

    Niepert, M.; Ahmed, M.; and Kutzkov, K. 2016. Learning Convolutional Neural Networks for Graphs. In International Conference on Machine Learning, 2014--2023. PMLR

    2016

  25. [33]

    Pan, H.; Bai, Y.; He, Z.; and Zhang, C. 2022. AAGCN: Adjacency-aware Graph Convolutional Network for Person Re-identification. Knowledge-Based Systems, 236: 107300

    2022

  26. [34]

    C.-C.; Lei, Y.; and Yang, B

    Pei, H.; Wei, B.; Chang, K. C.-C.; Lei, Y.; and Yang, B. 2020. Geom-GCN: Geometric Graph Convolutional Networks. arXiv preprint arXiv:2002.05287

    2020 arXiv

  27. [35]

    Peng, Z.; Huang, W.; Luo, M.; Zheng, Q.; Rong, Y.; Xu, T.; and Huang, J. 2020. Graph Representation Learning via Graphical Mutual Information Maximization. In Proceedings of the Web Conference 2020

    2020

  28. [36]

    Perozzi, B.; Al-Rfou, R.; and Skiena, S. 2014. Deepwalk: Online Learning of Social Representations. In Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 701--710

    2014

  29. [37]

    Qian, R.; Meng, T.; Gong, B.; Yang, M.-H.; Wang, H.; Belongie, S.; and Cui, Y. 2021. Spatiotemporal Contrastive Video Representation Learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    2021

  30. [38]

    Qin, L.; Welleck, S.; Khashabi, D.; and Choi, Y. 2022. COLD Decoding: Energy-based Constrained Text Generation with Langevin Dynamics. Advances in Neural Information Processing Systems, 35: 9538--9551

    2022

  31. [39]

    Roy, A.; Shu, J.; Elshocht, O.; Smeets, J.; Zhang, R.; and Li, P. 2023. GAD-EBM: Graph Anomaly Detection using Energy-Based Models. In NeurIPS 2023 Workshop: New Frontiers in Graph Learning

    2023

  32. [40]

    Sen, P.; Namata, G.; Bilgic, M.; Getoor, L.; Galligher, B.; and Eliassi-Rad, T. 2008. Collective Classification in Network Data. AI Magazine, 29(3): 93--93

    2008

  33. [41]

    Singh, A.; Chakraborty, O.; Varshney, A.; Panda, R.; Feris, R.; Saenko, K.; and Das, A. 2021. Semi-supervised Action Recognition with Temporal Contrastive Learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    2021

  34. [42]

    Srinivasan, B.; and Ribeiro, B. 2019. On the Equivalence between Node Embeddings and Structural Graph Representations. arXiv preprint arXiv:1910.00452

    2019 arXiv

  35. [43]

    Suhail, M.; Mittal, A.; Siddiquie, B.; Broaddus, C.; Eledath, J.; Medioni, G.; and Sigal, L. 2021. Energy-based Learning for Scene Graph Generation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    2021

  36. [44]

    Sun, Q.; Li, J.; Peng, H.; Wu, J.; Fu, X.; Ji, C.; and Philip, S. Y. 2022. Graph Structure Learning with Variational Information Bottleneck. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 36, 4165--4174

    2022

  37. [45]

    Szegedy, C.; Zaremba, W.; Sutskever, I.; Bruna, J.; Erhan, D.; Goodfellow, I.; and Fergus, R. 2013. Intriguing Properties of Neural Networks. arXiv preprint arXiv:1312.6199

    2013 arXiv

  38. [46]

    Ta, X.; Liu, Z.; Hu, X.; Yu, L.; Sun, L.; and Du, B. 2022. Adaptive Spatio-temporal Graph Neural Network for Traffic Forecasting. Knowledge-Based Systems, 242: 108199

    2022

  39. [47]

    Tang, J.; Sun, J.; Wang, C.; and Yang, Z. 2009. Social Influence Analysis in Large-scale Networks. In Proceedings of the 15th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

    2009

  40. [48]

    Veli c kovi \'c , P.; Cucurull, G.; Casanova, A.; Romero, A.; Lio, P.; and Bengio, Y. 2017. Graph Attention Networks. arXiv preprint arXiv:1710.10903

    2017 arXiv

  41. [49]

    L.; Li \`o , P.; Bengio, Y.; and Hjelm, R

    Veli c kovi \'c , P.; Fedus, W.; Hamilton, W. L.; Li \`o , P.; Bengio, Y.; and Hjelm, R. D. 2019. Deep Graph Infomax. In International Conference on Learning Representations

    2019

  42. [50]

    Wang, C.; Pan, S.; Hu, R.; Long, G.; Jiang, J.; and Zhang, C. 2019 a . Attributed Graph Clustering: A Deep Attentional Embedding Approach. arXiv preprint arXiv:1906.06532

    2019 arXiv

  43. [51]

    Wang, H.; Guo, X.; Deng, Z.-H.; and Lu, Y. 2022 a . Rethinking Minimal Sufficient Representation in Contrastive Learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 16041--16050

    2022

  44. [52]

    Wang, H.; Wang, J.; Wang, J.; Zhao, M.; Zhang, W.; Zhang, F.; Xie, X.; and Guo, M. 2018 a . GraphGAN: Graph Representation Learning with Generative Adversarial Nets. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 32

    2018

  45. [53]

    Wang, H.; Zhang, F.; Hou, M.; Xie, X.; Guo, M.; and Liu, Q. 2018 b . Shine: Signed Heterogeneous Information Network Embedding for Sentiment Link Prediction. In Proceedings of the Eleventh ACM International Conference on Web Search and Data Mining, 592--600

    2018

  46. [54]

    Wang, R.; Mou, S.; Wang, X.; Xiao, W.; Ju, Q.; Shi, C.; and Xie, X. 2021 a . Graph Structure Estimation Neural Networks. In Proceedings of the Web Conference 2021

    2021

  47. [55]

    Z.; and Orgun, M

    Wang, S.; Hu, L.; Wang, Y.; Cao, L.; Sheng, Q. Z.; and Orgun, M. 2019 b . Sequential Recommender Systems: Challenges, Progress and Prospects. arXiv preprint arXiv:2001.04830

    2019 arXiv

  48. [56]

    Wang, Y.; Min, Y.; Chen, X.; and Wu, J. 2021 b . Multi-view Graph Contrastive Representation Learning for Drug-Drug Interaction Prediction. In Proceedings of the Web Conference 2021

    2021

  49. [57]

    Wang, Y.; Wang, Y.; Yang, J.; and Lin, Z. 2022 b . A Unified Contrastive Energy-based Model for Understanding the Generative Ability of Adversarial Training. arXiv preprint arXiv:2203.13455

    2022 arXiv

  50. [58]

    Welling, M.; and Teh, Y. W. 2011. Bayesian Learning via Stochastic Gradient Langevin Dynamics. In Proceedings of the 28th International Conference on Machine Learning (ICML-11)

    2011

  51. [59]

    P.; and Yan, J

    Wu, Q.; Zhao, W.; Li, Z.; Wipf, D. P.; and Yan, J. 2022. Nodeformer: A Scalable Graph Structure Learning Transformer for Node Classification. Advances in Neural Information Processing Systems, 35: 27387--27401

    2022

  52. [60]

    Xinyi, Z.; and Chen, L. 2019. Capsule Graph Neural Network. In International Conference on Learning Representations

    2019

  53. [61]

    Xu, A.; Ren, X.; and Jia, R. 2022. Contrastive Novelty-Augmented Learning: Anticipating Outliers with Large Language Models. arXiv preprint arXiv:2211.15718

    2022 arXiv

  54. [62]

    Xu, K.; Hu, W.; Leskovec, J.; and Jegelka, S. 2018. How Powerful are Graph Neural Networks? arXiv preprint arXiv:1810.00826

    2018 arXiv

  55. [63]

    Yanardag, P.; and Vishwanathan, S. 2015. Deep Graph Kernels. In Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 1365--1374

    2015

  56. [64]

    Yang, X.; and Ji, S. 2023. M-EBM: Towards Understanding the Manifolds of Energy-Based Models. In Pacific-Asia Conference on Knowledge Discovery and Data Mining, 291--302. Springer

    2023

  57. [65]

    Yao, L.; Mao, C.; and Luo, Y. 2019. Graph Convolutional Networks for Text Classification. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 33

    2019

  58. [66]

    Yu, D.; Zhang, R.; Jiang, Z.; Wu, Y.; and Yang, Y. 2021 a . Graph-revised Convolutional Network. In Machine Learning and Knowledge Discovery in Databases: European Conference, ECML PKDD 2020, Ghent, Belgium, September 14--18, 2020, Proceedings, Part III. Springer

    2021

  59. [67]

    Yu, J.; Yin, H.; Li, J.; Wang, Q.; Hung, N. Q. V.; and Zhang, X. 2021 b . Self-Supervised Multi-Channel Hypergraph Convolutional Network for Social Recommendation. In Proceedings of The Web Conference 2021

    2021

  60. [68]

    Zhang, C.; Song, D.; Huang, C.; Swami, A.; and Chawla, N. V. 2019 a . Heterogeneous Graph Neural Network. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

    2019

  61. [69]

    Zhang, F.; Wang, R.; Zhou, F.; and Luo, Y. 2022. ERM: Energy-based Refined-attention Mechanism for Video Question Answering. IEEE Transactions on Circuits and Systems for Video Technology, 33(3): 1454--1467

    2022

  62. [70]

    Zhang, M.; Cui, Z.; Neumann, M.; and Chen, Y. 2018. An End-to-End Deep Learning Architecture for Graph Classification. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 32

    2018

  63. [71]

    Zhang, X.; Liu, H.; Li, Q.; and Wu, X.-M. 2019 b . Attributed Graph Clustering via Adaptive Graph Convolution. arXiv preprint arXiv:1906.01210

    2019 arXiv

  64. [72]

    Zhang, X.; and Zitnik, M. 2020. GNNG UARD : Defending Graph Neural Networks against Adversarial Attacks. Advances in Neural Information Processing Systems, 33: 9263--9275

    2020

  65. [73]

    Zhang, Z.; Bu, J.; Ester, M.; Zhang, J.; Yao, C.; Yu, Z.; and Wang, C. 2019 c . Hierarchical Graph Pooling with Structure Learning. arXiv preprint arXiv:1911.05954

    2019 arXiv

  66. [74]

    Zhao, J.; Wen, Q.; Ju, M.; Zhang, C.; and Ye, Y. 2023. Self-Supervised Graph Structure Refinement for Graph Neural Networks. In Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining

    2023

  67. [75]

    Zhao, T.; Liu, Y.; Neves, L.; Woodford, O.; Jiang, M.; and Shah, N. 2021. Data Augmentation for Graph Neural Networks. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35

    2021

  68. [76]

    Zheng, S.; Zhu, Z.; Zhang, X.; Liu, Z.; Cheng, J.; and Zhao, Y. 2020. Distribution-induced Bidirectional Generative Adversarial Network for Graph Representation Learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 7224--7233

    2020

  69. [77]

    Zhu, D.; Zhang, Z.; Cui, P.; and Zhu, W. 2019. Robust Graph Convolutional Networks against Adversarial Attacks. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

    2019

  70. [78]

    Zhu, Y.; Xu, W.; Zhang, J.; Du, Y.; Zhang, J.; Liu, Q.; Yang, C.; and Wu, S. 2021. A Survey on Graph Structure Learning: Progress and Opportunities. arXiv preprint arXiv:2103.03036

    2021 arXiv

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.