Pith. sign in

REVIEW 4 major objections 5 minor 62 references

CLDG: Contrastive Learning on Dynamic Graphs

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The paper argues that the temporal dimension of a dynamic graph can supply contrastive learning with positive pairs, replacing perturbation-based augmentation.

desk verdict Clean, simple dynamic-graph contrastive method with real engineering value, but the SOTA claim is not yet verifiable because the evaluation protocol is under-specified and Table III contradicts its own text. read the letter →

arxiv 2412.14451 v1 pith:52TCMRW6 submitted 2024-12-19 cs.LG cs.AI

Yiming Xu , Bin Shi , Teng Ma , Bo Dong , Haoyi Zhou , Qinghua Zheng This is my paper · ORCID
classification cs.LGcs.AI
keywords dynamicgraphcontrastivelearningtemporaltranslationinvariancerepresentationunsupervisedtimespanviewsamplingneuralnetworksnodeclassification
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

The paper claims that dynamic graphs carry a property it calls temporal translation invariance: the semantics and labels of a node tend to stay similar across different time windows, regardless of which encoder is used. CLDG exploits this by sampling several timespan views of the same dynamic graph and treating the same node, or a node and its neighborhood, across those views as positive contrastive pairs instead of corrupting the graph to make views. If this is right, temporal information itself supplies the self-supervised signal, avoiding the semantic damage that perturbation-based augmentations cause. On seven datasets CLDG reports state-of-the-art accuracy among twelve unsupervised baselines and exceeds supervised methods on four datasets, while using about 2,000 times fewer parameters and 130 times less training time than existing dynamic graph models.

What carries the argument

The load-bearing object is the timespan view sampling layer, which converts one dynamic graph into several temporally ordered views by choosing centers on the overall time interval and keeping edges within a window of width $\Delta t / s$, controlled by a view size factor $s$ and a view count $v$. The invariance assumption it rests on is temporal translation invariance: the same node in different windows is semantically the same entity, so its representations can be pulled together. Four sampling strategies are compared: sequential, high overlap, low overlap, and random. The counterintuitive result is that high overlap, which physically shares the most context, performs worst, while sequential and random sampling perform best. The method then uses InfoNCE to enforce local temporal translation invariance between node embeddings and global temporal translation invariance between a node and its neighborhood embeddings.

What would settle it

Take a dataset where node labels demonstrably change over time, such as a citation network re-labeled as researchers change fields or a synthetic temporal graph with periodic label flips. If CLDG's accuracy degrades toward or below a static contrastive baseline as the label-change rate rises, the temporal translation invariance assumption is the cause. A direct check is to compute same-node label agreement between non-overlapping windows and test whether it predicts CLDG's gain over perturbation-based baselines.

Watch

Extended reading notes

Core claim

CLDG's central discovery is that the temporal dimension of a dynamic graph is not just a feature to be modeled with recurrent networks but a source of free contrastive views. The authors first demonstrate empirically that nodes' predicted labels are similar across different timespans on their seven datasets, naming this temporal translation invariance. The method then samples $v$ timespan views via a sampling layer, encodes them with a shared-weight GCN, pools neighborhoods with a readout, projects through an MLP, and trains InfoNCE losses that pull the same node's representations across views together while pushing different nodes apart. Both a local variant (node-to-node) and a global variant (node-to-neighborhood) are reported, and the local variant is usually stronger. The reported effect is that an unsupervised model with no labels, no perturbation augmentations, and no sequence model reaches the best unsupervised classification results on all seven datasets and beats the best supervised baseline on four datasets.

Load-bearing premise

The premise is that a node's meaning and label stay stable across the time windows used as views, a property the paper observes empirically but does not quantify; the authors themselves note it fails when graph changes are non-continuous and labels constantly change.

Editorial extensions

If this is right

  • Dynamic graph representation learning can be done without recurrent networks: temporal cues are exploited implicitly by the contrastive objective, so model size and training time scale with the encoder rather than with a sequence model.
  • Existing static-graph contrastive methods can be adapted to dynamic graphs by replacing perturbation-based view generation with timespan view sampling, avoiding augmentation-induced label changes.
  • The encoder is swappable: GCN, GAT, and GraphSAGE all give competitive results inside CLDG, so future encoders can be integrated without redesigning the temporal component.
  • Low-overlap or non-overlapping timespan views are preferable to highly overlapping ones, suggesting that the model improves when the positive-pair task requires semantic rather than physical continuity.
  • The same sampling layer handles both discrete-time and continuous-time dynamic graphs, giving the method broader coverage than dynamic graph models that address only one formalism.

Reading between the lines

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

  • A direct testable extension is to measure temporal translation invariance on a new dataset before applying CLDG: if label agreement of the same node across non-overlapping windows is low, the positive-pair premise fails and the method should not be expected to help.
  • The high-overlap result hints that contrastive difficulty, not temporal proximity, is what drives representation quality, which could connect CLDG to curriculum learning and hard-negative selection.
  • In settings with gradual concept drift, one could anneal the weight of distant timespan views or resample windows per epoch so that positives respect a drift budget while keeping negatives informative.
  • Because the method only needs node identities to align views, it could be combined with temporal point-process encoders to add future-event prediction without giving up the lightweight contrastive objective.
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

4 major / 5 minor

Summary. The paper proposes CLDG, an unsupervised contrastive learning framework for dynamic graphs. The method assumes that node semantics and labels are approximately stable across different time intervals ('temporal translation invariance'), samples multiple timespan views of a dynamic graph, and trains a shared encoder with InfoNCE-style local and global contrastive losses. The authors report experiments on seven dynamic graph datasets and twelve baselines, claiming state-of-the-art unsupervised performance and competitiveness with supervised methods, while reducing model parameters and training time by large factors relative to existing dynamic-graph models.

Significance. If the empirical claims hold, CLDG would be a simple, lightweight, encoder-agnostic contrastive approach that unifies discrete-time and continuous-time dynamic graph representation learning, with strong practical appeal because it avoids sequence models and reconstructing future states. The paper includes extensive experiments, an ablation of sampling strategies, and a public code/data repository, which are valuable. However, the current evidence does not yet establish the central SOTA claim: the evaluation protocol is under-specified, some reported numbers contradict the text, and the core temporal translation invariance assumption is asserted rather than quantitatively validated.

major comments (4)
  1. [V-A3] The evaluation protocol that produces Table III is under-specified in a way that could change the results. The paper states a 1:1:8 train/validation/test split but does not say whether the split is by node or by time, and it says only that after unsupervised training the authors 'freeze the model parameters and output the learned representations for all nodes.' It is not stated whether those embeddings come from the full temporal graph, from one sampled timespan view, or from an aggregation of views. Since CLDG training samples timespan views from the whole timeline (Section IV-B, Eqs. 1-5), a time-based split would make the test interval visible during pretraining, while a node-based split still exposes each test node's complete temporal neighborhood. Please specify the split direction, the exact embedding source used for linear probing, and confirm that no test-interval edges or nodes enter the unsupervised training phase.
  2. [V-B, Table III] The sentence 'both of which outperform other unsupervised methods' is false as written. In Table III, CAW achieves 59.85 Accuracy on Bitcoinotc versus CLDGnode's 59.17; DGI achieves 73.19 Weighted-F1 on BITalpha versus CLDGnode's 72.90; and MNCI achieves 65.26 Weighted-F1 on Reddit versus CLDGnode's 62.87. The later statement that eleven metrics are optimal among the unsupervised methods is consistent with the table, but the blanket claim that both CLDG variants outperform all unsupervised baselines is not. Please correct the claim and report the exact set of metrics in which CLDG is and is not optimal.
  3. [IV-A] The entire method rests on the temporal translation invariance assumption, but it is not quantitatively established. Section IV-A reports only a qualitative observation (Figure 1) and states that prediction labels of the same node tend to be similar in different timespans 'regardless of the encoder used,' without giving any numerical agreement measure across the seven datasets. Because this assumption directly justifies the positive pairs in Eqs. 9-12, and the conclusion admits the method may fail when graph changes are non-continuous or labels constantly change, please provide a quantitative study of label/representation stability across timespan pairs, per dataset, and characterize the regime in which the assumption breaks down.
  4. [V-A4 and IV-B] Table III does not report the timespan view sampling configuration used for the main results. The paper introduces four sampling strategies, the view timespan factor s, and the number of views v, and Section V-C shows that these choices can change accuracy by about 1.4% on DBLP and 4.0% on TAX. Without stating which strategy and which s and v values were used for each dataset in Table III, the reported gains are not reproducible and could be configuration-dependent. Please report the exact configuration (or a pre-specified selection rule) used for the headline results.
minor comments (5)
  1. [VI] The abstract reports an average parameter reduction of 2,001.86 times, while the conclusion reports 'up to 2,000.86 times'; Table V shows per-dataset ratios that vary substantially. Please clarify whether the quoted number is an average or a maximum and define the averaging set.
  2. [V-C] The text refers to 'Table Va' and 'Table Vb' but the captions are labeled 'Table IV'; the table numbering should be made consistent.
  3. [IV-B] Equation (5) uses set notation for the interval containing |T_i - T_{i±1}| but does not specify how the tuple ordering is enforced in the random sampling strategy; please clarify the sampling procedure.
  4. [IV-B] For the high and low overlap strategies, the stated ranges for T1 (e.g., [min(T) + dt/(2s), max(T) - (2+v)·dt/(4s)]) implicitly assume a particular ordering and may be invalid for large v; the constraints should be stated precisely over the whole tuple.
  5. [V-B] The statement that CLDGnode 'outperforms the previous state-of-the-art GraphSAGE model by 1.47% on the average of all metrics' should specify whether this is averaged over datasets and both Accuracy and Weighted-F1, and should report the standard deviation or significance testing, as differences of this size may be within run-to-run variation.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the temporal-translation-invariance assumption is an observed empirical prior used as a label-free training signal, not a relabeled prediction of the paper's own outputs.

full rationale

The derivation chain in CLDG is self-contained: the paper observes temporal label stability empirically (Section IV-A) and then uses that observed property as an inductive bias by maximizing consistency of the same node's and node-neighborhood's representations across timespan views (Eqs. 9-12). The downstream node-classification evaluation is not used to construct the training signal; the contrastive loss is label-free, and the linear probe is trained only on the training split. No fitted parameter is renamed as a prediction, and no cited uniqueness theorem or prior author result is invoked to force the choice of temporal translation invariance. The only self-citation ([12], in an introductory list of risk-management applications) is an example and is not load-bearing. The paper's own limitation statement (Section VI) acknowledges that the assumption can fail, which further confirms that the claim is an empirical hypothesis rather than a definitional identity. Differences between the text and Table III (e.g., CAW 59.85 vs CLDGnode 59.17 on Bitcoinotc Accuracy) are correctness/reproducibility concerns, not circularity.

Assumptions & free parameters 3 free parameters · 3 assumptions · 0 invented entities

The central claim depends on the temporal translation invariance assumption and on hand-chosen hyperparameters (s, v, tau). No new physical or methodological entities are introduced.

free parameters (3)
  • temperature tau
    Inverse temperature in InfoNCE loss (Eq. 9-12); chosen by hand, not reported for main results.
  • view timespan factor s
    Controls the size of each timespan view (Eq. 1); sensitivity studied, but value used for Table III not specified.
  • number of views v
    Number of timespan views sampled; sensitivity studied, but value used for Table III not specified.
assumptions (3)
  • domain assumption Dynamic graphs exhibit temporal translation invariance: the same node's semantics and labels are similar across different timespans.
    Core inductive bias stated in Section IV-A; no quantitative empirical evidence is provided, and it fails when labels change over time (conclusion).
  • domain assumption The variation of the dynamic graph is continuous and smooth.
    Stated in the conclusion as a condition for applicability; used implicitly to justify that distant views share semantic context.
  • standard math InfoNCE provides a valid contrastive objective for this setup.
    Standard contrastive loss used as Eq. 9-12.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CLDG: Contrastive Learning on Dynamic Graphs." pith.science (2026). https://pith.science/paper/52TCMRW6

@misc{pith2026241214451,
  author       = {Pith},
  title        = {Pith review of: CLDG: Contrastive Learning on Dynamic Graphs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/52TCMRW6}},
  note         = {Machine review of arXiv:2412.14451}
}
read the original abstract

The graph with complex annotations is the most potent data type, whose constantly evolving motivates further exploration of the unsupervised dynamic graph representation. One of the representative paradigms is graph contrastive learning. It constructs self-supervised signals by maximizing the mutual information between the statistic graph's augmentation views. However, the semantics and labels may change within the augmentation process, causing a significant performance drop in downstream tasks. This drawback becomes greatly magnified on dynamic graphs. To address this problem, we designed a simple yet effective framework named CLDG. Firstly, we elaborate that dynamic graphs have temporal translation invariance at different levels. Then, we proposed a sampling layer to extract the temporally-persistent signals. It will encourage the node to maintain consistent local and global representations, i.e., temporal translation invariance under the timespan views. The extensive experiments demonstrate the effectiveness and efficiency of the method on seven datasets by outperforming eight unsupervised state-of-the-art baselines and showing competitiveness against four semi-supervised methods. Compared with the existing dynamic graph method, the number of model parameters and training time is reduced by an average of 2,001.86 times and 130.31 times on seven datasets, respectively.

Figures

Figures reproduced from arXiv: 2412.14451 by the authors.

Figure 1
Figure 1. Illustration of our basic idea. In the dataset used in this paper (details [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The architecture of the CLDG. The core of CLDG is implemented by maintaining local or global temporal translation invariance. Given an input [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Four candidate timespan view sampling strategies of CLDG. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Projection head of CLDG. achieve state-of-the-art on graph classification tasks by de￾signing more efficient readout functions through differentiable pooling [55] or a learnable filter function [56]. However, the graph pooling operation is not the core of this work. Th…
Figure 5
Figure 5. Figure 5: Parameter sensitivity of CLDG. Effect of epoch, batch size, output dimension and layers on the node classification. [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

62 extracted references · 34 canonical work pages

  1. [1]

    Graph convolutional neural networks for web-scale rec- ommender systems,

    R. Ying, R. He, K. Chen, P. Eksombatchai, W. L. Hamilton, and J. Leskovec, “Graph convolutional neural networks for web-scale rec- ommender systems,” in SIGKDD, 2018, pp. 974–983

    work page 2018

  2. [2]

    Graph neural networks in recommender systems: a survey,

    S. Wu, F. Sun, W. Zhang, X. Xie, and B. Cui, “Graph neural networks in recommender systems: a survey,” ACM Computing Surveys (CSUR) , 2020

    2020

  3. [3]

    Meta-learning on heterogeneous information networks for cold-start recommendation,

    Y . Lu, Y . Fang, and C. Shi, “Meta-learning on heterogeneous information networks for cold-start recommendation,” in SIGKDD, 2020, pp. 1563– 1573

    work page 2020

  4. [4]

    Combinatorial optimization with graph convolutional networks and guided tree search,

    Z. Li, Q. Chen, and V . Koltun, “Combinatorial optimization with graph convolutional networks and guided tree search,” NeurIPS, vol. 31, 2018

    work page 2018

  5. [5]

    Chip placement with deep reinforcement learning,

    A. Mirhoseini, A. Goldie, M. Yazgan, J. Jiang, E. Songhori, S. Wang, Y .-J. Lee, E. Johnson, O. Pathak, S. Bae et al. , “Chip placement with deep reinforcement learning,” arXiv preprint arXiv:2004.10746 , 2020

    arXiv 2004

  6. [6]

    Combinatorial optimization and reasoning with graph neural networks,

    Q. Cappart, D. Ch ´etelat, E. Khalil, A. Lodi, C. Morris, and P. Veliˇckovi´c, “Combinatorial optimization and reasoning with graph neural networks,” arXiv preprint arXiv:2102.09544 , 2021

    arXiv 2021

  7. [7]

    Spatio-temporal graph convolutional networks: A deep learning framework for traffic forecasting,

    B. Yu, H. Yin, and Z. Zhu, “Spatio-temporal graph convolutional networks: A deep learning framework for traffic forecasting,” arXiv preprint arXiv:1709.04875, 2017

    arXiv 2017

  8. [8]

    Attention based spatial- temporal graph convolutional networks for traffic flow forecasting,

    S. Guo, Y . Lin, N. Feng, C. Song, and H. Wan, “Attention based spatial- temporal graph convolutional networks for traffic flow forecasting,” in Proceedings of the AAAI conference on artificial intelligence , vol. 33, no. 01, 2019, pp. 922–929

    work page 2019

Show all 62 references
  1. [9]

    Eta prediction with graph neural networks in google maps,

    A. Derrow-Pinion, J. She, D. Wong, O. Lange, T. Hester, L. Perez, M. Nunkesser, S. Lee, X. Guo, B. Wiltshire et al., “Eta prediction with graph neural networks in google maps,” in Proceedings of the 30th ACM International Conference on Information & Knowledge Management , 2021...

    2021

  2. [10]

    Cash-out user detection based on attributed heterogeneous information network with a hierarchical attention mechanism,

    B. Hu, Z. Zhang, C. Shi, J. Zhou, X. Li, and Y . Qi, “Cash-out user detection based on attributed heterogeneous information network with a hierarchical attention mechanism,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 33, no. 01, 2019, pp. 946–953

    2019

  3. [11]

    Financial defaulter detection on online credit payment via multi-view attributed heterogeneous information network,

    Q. Zhong, Y . Liu, X. Ao, B. Hu, J. Feng, J. Tang, and Q. He, “Financial defaulter detection on online credit payment via multi-view attributed heterogeneous information network,” in Proceedings of The Web Conference 2020, 2020, pp. 785–795

    2020

  4. [12]

    Tax evasion detection with fbne-pu algorithm based on pncgcn and pu learning,

    Y . Gao, B. Shi, B. Dong, Y . Wang, L. Mi, and Q. Zheng, “Tax evasion detection with fbne-pu algorithm based on pncgcn and pu learning,” IEEE Transactions on Knowledge and Data Engineering , 2021

    2021

  5. [13]

    Deep graph infomax

    P. Velickovic, W. Fedus, W. L. Hamilton, P. Li `o, Y . Bengio, and R. D. Hjelm, “Deep graph infomax.” ICLR (Poster), vol. 2, no. 3, p. 4, 2019

    2019

  6. [14]

    Learning deep representations by mutual information estimation and maximization,

    R. D. Hjelm, A. Fedorov, S. Lavoie-Marchildon, K. Grewal, P. Bach- man, A. Trischler, and Y . Bengio, “Learning deep representations by mutual information estimation and maximization,” arXiv preprint arXiv:1808.06670, 2018

    2018 arXiv

  7. [15]

    Graph contrastive learning with augmentations,

    Y . You, T. Chen, Y . Sui, T. Chen, Z. Wang, and Y . Shen, “Graph contrastive learning with augmentations,” NeurIPS, vol. 33, pp. 5812– 5823, 2020

    2020

  8. [16]

    Graph contrastive learning with adaptive augmentation,

    Y . Zhu, Y . Xu, F. Yu, Q. Liu, S. Wu, and L. Wang, “Graph contrastive learning with adaptive augmentation,” in Proceedings of the Web Con- ference 2021, 2021, pp. 2069–2080

    2021

  9. [17]

    Deep graph contrastive representation learning,

    ——, “Deep graph contrastive representation learning,” arXiv preprint arXiv:2006.04131, 2020

    2006 arXiv

  10. [18]

    Contrastive multi-view represen- tation learning on graphs,

    K. Hassani and A. H. Khasahmadi, “Contrastive multi-view represen- tation learning on graphs,” in International Conference on Machine Learning. PMLR, 2020, pp. 4116–4126

    2020

  11. [19]

    From canonical correlation analysis to self-supervised graph neural networks,

    H. Zhang, Q. Wu, J. Yan, D. Wipf, and P. S. Yu, “From canonical correlation analysis to self-supervised graph neural networks,” NeurIPS, vol. 34, 2021

    2021

  12. [20]

    Semi-supervised classification with graph convolutional networks,

    T. N. Kipf and M. Welling, “Semi-supervised classification with graph convolutional networks,” arXiv preprint arXiv:1609.02907 , 2016

    2016 arXiv

  13. [21]

    Graph attention networks,

    P. Veli ˇckovi´c, G. Cucurull, A. Casanova, A. Romero, P. Lio, and Y . Ben- gio, “Graph attention networks,” arXiv preprint arXiv:1710.10903, 2017

    2017 arXiv

  14. [22]

    A logical calculus of the ideas immanent in nervous activity,

    W. S. McCulloch and W. Pitts, “A logical calculus of the ideas immanent in nervous activity,” The bulletin of mathematical biophysics , vol. 5, no. 4, pp. 115–133, 1943

    1943

  15. [23]

    Simgrace: A simple framework for graph contrastive learning without data augmentation,

    J. Xia, L. Wu, J. Chen, B. Hu, and S. Z. Li, “Simgrace: A simple framework for graph contrastive learning without data augmentation,” in Proceedings of the ACM Web Conference 2022 , 2022, pp. 1070–1079

    2022

  16. [24]

    Graphmix: Improved training of gnns for semi-supervised learning,

    V . Verma, M. Qu, K. Kawaguchi, A. Lamb, Y . Bengio, J. Kannala, and J. Tang, “Graphmix: Improved training of gnns for semi-supervised learning,” arXiv preprint arXiv:1909.11715 , 2019

    1909 arXiv

  17. [25]

    Graph random neural networks for semi- supervised learning on graphs,

    W. Feng, J. Zhang, Y . Dong, Y . Han, H. Luan, Q. Xu, Q. Yang, E. Kharlamov, and J. Tang, “Graph random neural networks for semi- supervised learning on graphs,” NeurIPS, vol. 33, pp. 22 092–22 103, 2020

    2020

  18. [26]

    Graph stochas- tic neural networks for semi-supervised learning,

    H. Wang, C. Zhou, X. Chen, J. Wu, S. Pan, and J. Wang, “Graph stochas- tic neural networks for semi-supervised learning,” NeurIPS, vol. 33, pp. 19 839–19 848, 2020

    2020

  19. [27]

    Unsupervised feature learning via non-parametric instance discrimination,

    Z. Wu, Y . Xiong, S. X. Yu, and D. Lin, “Unsupervised feature learning via non-parametric instance discrimination,” in CVPR, 2018, pp. 3733– 3742

    2018

  20. [28]

    Representation learning with contrastive predictive coding,

    A. Van den Oord, Y . Li, and O. Vinyals, “Representation learning with contrastive predictive coding,” arXiv e-prints, pp. arXiv–1807, 2018

    2018

  21. [29]

    Momentum contrast for unsupervised visual representation learning,

    K. He, H. Fan, Y . Wu, S. Xie, and R. Girshick, “Momentum contrast for unsupervised visual representation learning,” in CVPR, 2020, pp. 9729–9738

    2020

  22. [30]

    A simple framework for contrastive learning of visual representations,

    T. Chen, S. Kornblith, M. Norouzi, and G. Hinton, “A simple framework for contrastive learning of visual representations,” in International conference on machine learning . PMLR, 2020, pp. 1597–1607

    2020

  23. [31]

    Bootstrap your own latent-a new approach to self-supervised learning,

    J.-B. Grill, F. Strub, F. Altch ´e, C. Tallec, P. Richemond, E. Buchatskaya, C. Doersch, B. Avila Pires, Z. Guo, M. Gheshlaghi Azar et al. , “Bootstrap your own latent-a new approach to self-supervised learning,” NeurIPS, vol. 33, pp. 21 271–21 284, 2020

    2020

  24. [32]

    Simcse: Simple contrastive learning of sentence embeddings,

    T. Gao, X. Yao, and D. Chen, “Simcse: Simple contrastive learning of sentence embeddings,” arXiv preprint arXiv:2104.08821 , 2021

    2021 arXiv

  25. [33]

    Diffusion improves graph learning,

    J. Klicpera, S. Weißenberger, and S. G ¨unnemann, “Diffusion improves graph learning,” NeurIPS, vol. 32, 2019

    2019

  26. [34]

    Line: Large-scale information network embedding,

    J. Tang, M. Qu, M. Wang, M. Zhang, J. Yan, and Q. Mei, “Line: Large-scale information network embedding,” in Proceedings of the 24th international conference on world wide web , 2015, pp. 1067–1077

    2015

  27. [35]

    Deepwalk: Online learning of social representations,

    B. Perozzi, R. Al-Rfou, and S. Skiena, “Deepwalk: Online learning of social representations,” in SIGKDD, 2014, pp. 701–710

    2014

  28. [36]

    Combining label propagation and simple models out-performs graph neural net- works,

    Q. Huang, H. He, A. Singh, S.-N. Lim, and A. R. Benson, “Combining label propagation and simple models out-performs graph neural net- works,” arXiv preprint arXiv:2010.13993 , 2020

    2010 arXiv

  29. [37]

    Node feature extraction by self-supervised multi-scale neighborhood prediction,

    E. Chien, W.-C. Chang, C.-J. Hsieh, H.-F. Yu, J. Zhang, O. Milenkovic, and I. S. Dhillon, “Node feature extraction by self-supervised multi-scale neighborhood prediction,” arXiv preprint arXiv:2111.00064 , 2021

    2021 arXiv

  30. [38]

    Dyngem: Deep embedding method for dynamic graphs,

    P. Goyal, N. Kamra, X. He, and Y . Liu, “Dyngem: Deep embedding method for dynamic graphs,” arXiv preprint arXiv:1805.11273 , 2018

    2018 arXiv

  31. [39]

    Dysat: Deep neural representation learning on dynamic graphs via self-attention networks,

    A. Sankar, Y . Wu, L. Gou, W. Zhang, and H. Yang, “Dysat: Deep neural representation learning on dynamic graphs via self-attention networks,” in Proceedings of the 13th International Conference on Web Search and Data Mining, 2020, pp. 519–527

    2020

  32. [40]

    In- ductive representation learning on temporal graphs,

    D. Xu, C. Ruan, E. Korpeoglu, S. Kumar, and K. Achan, “In- ductive representation learning on temporal graphs,” arXiv preprint arXiv:2002.07962, 2020

    2002 arXiv

  33. [41]

    Evolvegcn: Evolving graph convolutional networks for dynamic graphs,

    A. Pareja, G. Domeniconi, J. Chen, T. Ma, T. Suzumura, H. Kanezashi, T. Kaler, T. Schardl, and C. Leiserson, “Evolvegcn: Evolving graph convolutional networks for dynamic graphs,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 34, no. 04, 2020, pp. 5363– 5370

    2020

  34. [42]

    Modeling dynamic heterogeneous network for link prediction using hierarchical attention with temporal rnn,

    H. Xue, L. Yang, W. Jiang, Y . Wei, Y . Hu, and Y . Lin, “Modeling dynamic heterogeneous network for link prediction using hierarchical attention with temporal rnn,” in Joint European Conference on Machine Learning and Knowledge Discovery in Databases . Springer, 2020, pp. 282–298

    2020

  35. [43]

    Temporal-aware graph neural network for credit risk prediction,

    D. Wang, Z. Zhang, J. Zhou, P. Cui, J. Fang, Q. Jia, Y . Fang, and Y . Qi, “Temporal-aware graph neural network for credit risk prediction,” in Proceedings of the 2021 SIAM International Conference on Data Mining (SDM). SIAM, 2021, pp. 702–710

    2021

  36. [44]

    Embedding temporal network via neighborhood formation,

    Y . Zuo, G. Liu, H. Lin, J. Guo, X. Hu, and J. Wu, “Embedding temporal network via neighborhood formation,” in SIGKDD, 2018, pp. 2857– 2866

    2018

  37. [45]

    Predicting dynamic embedding trajectory in temporal interaction networks,

    S. Kumar, X. Zhang, and J. Leskovec, “Predicting dynamic embedding trajectory in temporal interaction networks,” in SIGKDD, 2019, pp. 1269–1278

    2019

  38. [46]

    Dyrep: Learning representations over dynamic graphs,

    R. Trivedi, M. Farajtabar, P. Biswal, and H. Zha, “Dyrep: Learning representations over dynamic graphs,” in ICLR, 2019

    2019

  39. [47]

    Temporal network embedding with micro-and macro-dynamics,

    Y . Lu, X. Wang, C. Shi, P. S. Yu, and Y . Ye, “Temporal network embedding with micro-and macro-dynamics,” in Proceedings of the 28th ACM international conference on information and knowledge management, 2019, pp. 469–478

    2019

  40. [48]

    Spatio-temporal attentive rnn for node classification in temporal attributed graphs

    D. Xu, W. Cheng, D. Luo, X. Liu, and X. Zhang, “Spatio-temporal attentive rnn for node classification in temporal attributed graphs.” in IJCAI, 2019, pp. 3947–3953

    2019

  41. [49]

    dyngraph2vec: Captur- ing network dynamics using dynamic graph representation learning,

    P. Goyal, S. R. Chhetri, and A. Canedo, “dyngraph2vec: Captur- ing network dynamics using dynamic graph representation learning,” Knowledge-Based Systems, vol. 187, p. 104816, 2020

    2020

  42. [50]

    Autoencoders, unsupervised learning, and deep architectures,

    P. Baldi, “Autoencoders, unsupervised learning, and deep architectures,” in Proceedings of ICML workshop on unsupervised and transfer learn- ing. JMLR Workshop and Conference Proceedings, 2012, pp. 37–49

    2012

  43. [51]

    Long short-term memory,

    S. Hochreiter and J. Schmidhuber, “Long short-term memory,” Neural computation, vol. 9, no. 8, pp. 1735–1780, 1997

    1997

  44. [52]

    Learning phrase representations using rnn encoder-decoder for statistical machine translation,

    K. Cho, B. Van Merri ¨enboer, C. Gulcehre, D. Bahdanau, F. Bougares, H. Schwenk, and Y . Bengio, “Learning phrase representations using rnn encoder-decoder for statistical machine translation,” arXiv preprint arXiv:1406.1078, 2014

    2014 arXiv

  45. [53]

    Dynamic network embedding survey,

    G. Xue, M. Zhong, J. Li, J. Chen, C. Zhai, and R. Kong, “Dynamic network embedding survey,” Neurocomputing, vol. 472, pp. 212–223, 2022

    2022

  46. [54]

    Representation learning for dynamic graphs: A survey

    S. M. Kazemi, R. Goel, K. Jain, I. Kobyzev, A. Sethi, P. Forsyth, and P. Poupart, “Representation learning for dynamic graphs: A survey.” J. Mach. Learn. Res. , vol. 21, no. 70, pp. 1–73, 2020

    2020

  47. [55]

    Hierarchical graph representation learning with differentiable pooling,

    Z. Ying, J. You, C. Morris, X. Ren, W. Hamilton, and J. Leskovec, “Hierarchical graph representation learning with differentiable pooling,” NeurIPS, vol. 31, 2018

    2018

  48. [56]

    Graph filtration learning,

    C. Hofer, F. Graf, B. Rieck, M. Niethammer, and R. Kwitt, “Graph filtration learning,” in International Conference on Machine Learning . PMLR, 2020, pp. 4314–4323

    2020

  49. [57]

    Edge weight prediction in weighted signed networks,

    S. Kumar, F. Spezzano, V . Subrahmanian, and C. Faloutsos, “Edge weight prediction in weighted signed networks,” in 2016 IEEE 16th International Conference on Data Mining (ICDM) . IEEE, 2016, pp. 221–230

    2016

  50. [58]

    Inductive representation learning in temporal networks via mining neighborhood and community influences,

    M. Liu and Y . Liu, “Inductive representation learning in temporal networks via mining neighborhood and community influences,” in Pro- ceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval , 2021, pp. 2202–2206

    2021

  51. [59]

    Learning from labeled and unlabeled data with label propagation,

    X. Zhu and Z. Ghahramani, “Learning from labeled and unlabeled data with label propagation,” 2002

    2002

  52. [60]

    Inductive representation learning on large graphs,

    W. Hamilton, Z. Ying, and J. Leskovec, “Inductive representation learning on large graphs,” NeurIPS, vol. 30, 2017

    2017

  53. [61]

    Inductive representation learning in temporal networks via causal anonymous walks,

    Y . Wang, Y .-Y . Chang, Y . Liu, J. Leskovec, and P. Li, “Inductive representation learning in temporal networks via causal anonymous walks,” arXiv preprint arXiv:2101.05974 , 2021

    2021 arXiv

  54. [62]

    Adam: A method for stochastic optimization,

    D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv preprint arXiv:1412.6980 , 2014

    2014 arXiv

Pith tools

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