Pith. sign in

REVIEW 4 major objections 5 minor 51 references

Understanding GNNs and Homophily in Dynamic Node Classification

T0 review · 4 major / 5 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 In dynamic graphs, GCN accuracy is governed by future-label agreement with current neighbors, not static homophily.

desk verdict The dynamic homophily idea is worth engaging with, but the central theorem only holds for one layer; the multi-layer generalization is invalid, and that breaks the theoretical story. read the letter →

arxiv 2504.20421 v1 pith:PSDATVVC submitted 2025-04-29 cs.LG

Michael Ito , Danai Koutra , Jenna Wiens This is my paper · ORCID
classification cs.LG
keywords graphneuralnetworkshomophilydynamicgraphsnodeclassificationGCNtemporalAUROCmessagepassing
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

Dynamic node classification asks a GNN to predict next-step labels from the current graph. The paper's central claim is that a GCN's discriminative power at time $t$ is governed by the probability that a node's future label matches its neighbors' current labels, encoded as class-wise probabilities $h^+_t$ and $h^-_t$. Formally, the expected representation gap between future positive and negative nodes after $l$ GCN layers is $2\mu_t(h^+_t + h^-_t - 1)^l$, and the expected AUROC is bounded above by a function of the same probabilities. The paper names this quantity dynamic homophily and shows empirically that it tracks GNN AUROC across epidemiological, social, and protein-interaction graphs, while static homophily stays high and does not track performance. If the claim is right, static homophily is the wrong diagnostic for evolving graphs, and low dynamic homophily, not low static homophily, is what breaks GNNs.

What carries the argument

The machinery is the single-layer mean-aggregation calculation that partitions a node's neighbors by whether their current label equals the node's future label. For one GCN layer, a future positive node's expected representation becomes $(2h^+_t - 1)\mu_t$ and a future negative node's becomes $(1 - 2h^-_t)\mu_t$, so the class gap is $2\mu_t(h^+_t + h^-_t - 1)$; induction over layers raises the factor to the $l$-th power, yielding Theorem 3.2. The same partition trick feeds Theorem 3.4, which bounds expected AUROC by $1 - \Phi\left(-2\mu_t(h^+_t+h^-_t-1)^l / (v^+_{t+1}(l)+v^-_{t+1}(l))\right)$ with recursive variance lower bounds $v^+_{t+1}(l)=h^{+2}_t v^+_{t+1}(l-1)+(1-h^+_t)^2 v^-_{t+1}(l-1)$ and symmetrically for $v^-$. Gaussian Lipschitz concentration (Theorem 3.3) then guarantees the empirical class gap tracks the expected gap with high probability when class sizes are large. The named object that carries the argument is dynamic homophily, $h^D_t = \frac{1}{|C|}\sum_{c\in C} h^c_t$, the average probability that a node's label at $t+1$ equals its neighbor's label at $t$.

What would settle it

On a dynamic stochastic block model with controlled $h^+_t$ and $h^-_t$ and large node degrees, compute the empirical AUROC and true variance of a one-layer GCN; the central claim fails if AUROC does not move monotonically with $h^+_t + h^-_t - 1$, or if the true variance is smaller than the recursive lower bound $v^+_{t+1}(1)+v^-_{t+1}(1)$ used in Theorem 3.4.

Watch

Extended reading notes

Core claim

The discovery is stated exactly in Theorem 3.2: after $l$ layers of a GCN with mean aggregation, the expected representation gap between nodes that will be positive at $t+1$ and nodes that will be negative at $t+1$ is $2\mu_t(h^+_t + h^-_t - 1)^l$, where $h^+_t$ and $h^-_t$ are the probabilities that a future positive (resp. negative) node's label at $t+1$ equals its neighbor's label at $t$. This quantity interpolates between high dynamic homophily, where the gap grows with depth, and low dynamic homophily, where the gap shrinks or changes sign. Theorem 3.4 converts the gap into an upper bound on expected AUROC using recursively defined variance lower bounds that depend only on the same probabilities, which predicts that odd-layer GCNs improve monotonically in dynamic homophily while even layers can recover when both homophily levels are low. The empirical counterpart is that dynamic homophily achieves median Spearman correlations around 0.75 with homophilous GNN AUROC and 0.78 with heterophilous GNN AUROC across the tested dynamic graphs, while static homophily's medians are 0.26 and 0.38, respectively.

Load-bearing premise

The argument depends on replacing each node's neighbor aggregation with a weighted mixture of two Gaussian draws, one per class, weighted by $h^+_t$ and $1-h^+_t$ (and symmetrically for negatives), ignoring how many neighbors the node actually averages over and how degree normalization changes the sum.

Editorial extensions

If this is right

  • If Theorem 3.2 is correct, oversmoothing in dynamic GCNs is not a fixed depth effect: the class gap shrinks by the factor $h^+_t + h^-_t - 1$ per layer, so low dynamic homophily accelerates the loss of discriminative signal with depth.
  • Even-layer GCNs become a partial remedy in low dynamic homophily regimes: when both class probabilities are below 0.5, the squared factor makes the expected gap positive again, so depth can restore some separability.
  • Dynamic homophily, not static homophily, is the quantity to measure in next-step node classification; the experiments show static homophily remains high while dynamic homophily and AUROC move together across time.
  • Heterophilous GNN designs (ego-neighbor separation, residual connections) improve absolute performance under low dynamic homophily but still degrade as dynamic homophily falls, so the limitation is not specific to plain GCNs.
  • Combining intermediate GCN layer representations is a concrete design direction for dynamic GNNs that must perform well across changing dynamic homophily levels.

Reading between the lines

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

  • Beyond the paper, one can test whether the same separation formula holds for other mean-aggregation message-passing architectures by replacing the linear-GCN induction with their propagation rules; the paper only analyzes linear GCNs, so this is an extension.
  • The variance bound in Theorem 3.4 ignores the actual count of neighbors and degree normalization by modeling each aggregate as two Gaussian draws; a tighter analysis would keep the full neighbor sum in the variance recursion.
  • Dynamic homophily could serve as a dataset-level diagnostic for temporal graph benchmarks: graphs with high static homophily but low dynamic homophily should be exactly where standard GNNs underperform, which can be checked on existing temporal benchmark suites.
  • In multiclass settings, the appendix's dynamic compatibility matrix, rather than the scalar average, is the natural correlate of per-class AUROC; testing whether row distances of that matrix predict class-wise performance would extend the binary result.
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.

Referee Report

4 major / 5 minor

Summary. The paper studies GNN performance in dynamic node classification and proposes dynamic homophily, defined as the probability that a node's future label equals its neighbor's current label. It derives theoretical claims for linear GCNs: Theorem 3.2 states that after l layers the expected representation difference between future positive and negative nodes equals 2*mu_t*(h_t^+ + h_t^- - 1)^l; Theorem 3.3 gives a concentration bound for this difference; and Theorem 3.4 bounds expected AUROC via recursively defined variance lower bounds. It also defines multiclass and continuous-time extensions. Empirically, it compares dynamic homophily with static homophily as correlates of AUROC for seven GNN architectures on SI-model synthetic graphs, Higgs Twitter networks, and DPPIN protein interaction networks, reporting that dynamic homophily has substantially higher Spearman correlations with GNN AUROC than static homophily.

Significance. If the theoretical results were correct, the paper would provide a clean first characterization of GNN discriminative power in dynamic graphs and a practically useful metric. The empirical study is extensive and the core observation that static homophily can stay high while GNN performance collapses, whereas dynamic homophily tracks performance, is interesting and potentially valuable. However, the central theoretical claim, Theorem 3.2, is false for l >= 2, and the variance argument behind Theorem 3.4 is not a valid lower bound. Because the paper's main message is theory-driven, the empirical correlations do not repair the theoretical foundation. The dynamic homophily measure itself and the empirical findings may be salvageable as a standalone descriptive contribution, but the paper as submitted does not establish its central claims.

major comments (4)
  1. [Section 3.2, Theorem 3.2 and Appendix B.2] The induction step in the proof of Theorem 3.2 is invalid for l >= 2. In Eqs. (29)-(30), the proof writes the expected future-positive representation at layer l+1 in terms of expectations over current-labeled classes V_t^+ and V_t^-, and then applies the inductive hypothesis to E_{i in V_t^+}[h_t^(l)(i)] - E_{i in V_t^-}[h_t^(l)(i)]. But the inductive hypothesis is a statement about expectations over future-labeled nodes V_{t+1}^+ and V_{t+1}^-; the two quantities are not equal in general, since the l-layer separation of current-label classes is governed by static homophily rather than by the dynamic quantities h_t^+ and h_t^-. The proof also uses a single global factor h_t^+ for every node, although the expectation of a degree-normalized neighborhood average depends on the local neighbor composition and on the degree of each center. The theorem is false for l = 2. Consider the 4-node path with self-loops, y_0 = [+, +, -, -], y_1 = [-, -, +, +], x_t(i) = y_t(i), mu = 1, sigma -> 0. Using the node-averaged definition in Appendix D.4, h_t^+ = h_t^- = 1/6, so (h_t^+ + h_t^- - 1)^2 = 4/9 and Theorem 3.2 predicts a separation of 8/9. Direct computation gives h_t^(1) = [1, 1/3, -1/3, -1] and h_t^(2) = [2/3, 1/3, -1/3, -2/3]; the mean of the future-positive nodes {3,4} is -1/2 and the mean of the future-negative nodes {1,2} is +1/2, so the true separation is -1, not +8/9. The equality fails in both sign and magnitude, so the central theoretical claim is not established.
  2. [Section 3.2, Theorem 3.4 and Appendix B.4] The variance lower bounds in Eq. (12) are not valid for GCN mean aggregation. The proof treats each node's aggregated representation as a two-term weighted sum h_t^+ Z_+ + (1 - h_t^+) Z_- rather than as an average over the actual neighborhood. For a node with d neighbors and a same-class neighbor count S ~ Bin(d, h), the variance of the aggregation (1/d) sum_{j=1}^d Z_j is sigma^2/d + 4 mu^2 h(1-h)/d, which can be much smaller than h^2 sigma^2 + (1-h)^2 sigma^2 when d is large. Thus the recursively defined quantities v_{t+1}^+(l) and v_{t+1}^-(l) are not lower bounds on the true variances; in common high-degree settings they are larger than the true variance. Consequently the claimed upper bound on expected AUROC in Eq. (11) has the wrong direction and is not guaranteed. Since Theorem 3.4 underlies the layer-wise AUROC predictions in Figure 2 and the design recommendations, this is a load-bearing gap, not a minor technicality.
  3. [Section 3.1, Lemma 3.1, Eq. (4)] The AUROC characterization in Eq. (4) is not correct as written. The denominator should be the standard deviation of the difference h_t^(l)(i) - h_t^(l)(j) for a random positive-negative pair, but the displayed formula uses V_{i,j}[h_t^(l)(i) + h_t^(l)(j)] and omits the covariance between the two representations. The two nodes in a random positive-negative pair have correlated representations because they share neighbors after multiple GCN layers, and the expression also assumes that the difference is Gaussian with a single variance. No independence or homoscedasticity assumption is stated in the lemma. Theorem 3.4 substitutes its variance recursion into this formula, so the error propagates to the main AUROC bound.
  4. [Section 3.2, Theorem 3.3 and Appendix B.3] The concentration bound in Theorem 3.3 is not proven for general graphs. In Eq. (41), the proof claims that x_t(i) appears in the positive-class mean 'precisely d_t(i)*l times' and uses this to derive the Lipschitz norm L_{t,+}^(l) in Eq. (9). This count is not correct for arbitrary graph topologies: the coefficient of x_t(i) in h_t^(l)(j) is a sum over length-l walks from j to i of products of inverse degrees, and it cannot be reduced to a fixed factor d_t(i)^l independent of the walk structure. The explicit bounds in Eqs. (9)-(10) are therefore unsupported. The concentration statement may be repairable by a more careful Lipschitz analysis, but as submitted the proof does not establish the theorem as stated.
minor comments (5)
  1. [Appendix B.2, Eqs. (27)-(28)] The expectation subscripts in Eqs. (27) and (28) are garbled: both lines use E_{j|V_t^+}[x_t(j)] where the negative-case derivation requires one term over V_t^- and one over V_t^+, and the sign algebra leading to (2h_t^+ - 1)*mu is written incorrectly (the second term should be -(1-h_t^+)*mu rather than +(h_t^+ - 1)*mu).
  2. [Appendix D.4, Eq. (83)] Dynamic homophily is computed in the experiments as a node-averaged quantity, while Theorem 3.2 defines h_t^+ and h_t^- as probabilities over neighbor draws. These two notions agree only when every future node is weighted equally and neighbors are sampled uniformly; the paper should state this equivalence explicitly and confirm that the implementation matches Definition 3.5.
  3. [Appendix B.2-B.3] The propagation rule in Eq. (2) uses the degree d(i)+1 after adding a self-loop, but the proof recursions in Eqs. (24), (39), and (41) switch between d_t(i)+1 and d_t(i)^l without explanation, and the notation hat{N}_t(i)^l is never defined.
  4. [Checklist item 3(a)] The checklist states 'No' for code, data, and instructions needed to reproduce the main experimental results, although the footnote in Section 4.3 provides a GitHub URL. This discrepancy should be resolved.
  5. [Figure 2 and surrounding text] The caption and text should make clear that Figure 2 visualizes the upper bound from Eq. (11) rather than measured AUROC, especially because the bound itself is based on the variance recursion whose validity is questioned above.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: dynamic homophily is independently defined from labels and edges, and the empirical correlation with AUROC is a falsifiable consistency check rather than a fitted identity.

full rationale

The paper's theoretical chain is self-contained: h+_t and h-_t (Eqs. 6-7) are defined directly from node labels and neighbor sets, not from GCN representations or from the performance measures being predicted. Theorem 3.2 derives the expected l-layer separation from the GCN mean-aggregation rule and the Gaussian feature assumption; for l=1 the derivation is a direct expansion, and the induction is claimed rather than imported from prior work. The proposed dynamic homophily (Definition 3.5) is an affine average of these same h values, so the later finding that hD correlates with AUROC is an empirical validation of the theorem's stated dependence, not a parameter fitted to AUROC and renamed as a prediction; it can fail, and the paper reports weak correlations (e.g., SAGE on Math). No load-bearing self-citation or imported uniqueness theorem appears in the derivation. Concerns about the validity of the induction for l>=2 and the variance approximation in Theorem 3.4 are correctness/robustness issues, not circularity: the claimed equations still reference independently defined quantities rather than reducing to their own inputs. The absence of code/data in the checklist is a reproducibility issue, not circularity. Under the required standard, no circular step is exhibited, so the appropriate score is 0.

Assumptions & free parameters 0 free parameters · 3 assumptions · 2 invented entities

The central theoretical claim rests on Gaussian independent features and linear GCN aggregation; both are standard in the static homophily literature. The variance computation in Theorem 3.4 adds an implicit assumption that the aggregated representation can be treated as a two-component mixture, which is not justified and affects the bound's validity. Dynamic homophily is a new construct, but it is defined from quantities already present in the derived expressions.

assumptions (3)
  • domain assumption Node features at each time are Gaussian with mean y_t(i)·μ_t and variance σ_t², and are independent across nodes.
    Stated in Section 3.1; used in Theorem 3.2, Theorem 3.3, and Theorem 3.4 proofs. Independence is implicit in the variance computations in Appendix B.4.
  • domain assumption The GCN under analysis is linear with mean aggregation (no nonlinearities).
    Stated in Section 3.1; the theory applies to linear GCNs, while experiments generalize to nonlinear GNNs.
  • standard math Lipschitz concentration inequality for Gaussian functions (Wainwright Lemma B.5) and the AUROC concentration bound (Agarwal et al.), used as black boxes.
    Cited and used in Appendix B.3 and B.4.
invented entities (2)
  • Dynamic homophily h^D_t
    purpose: A measure of the probability that a node's future label matches its neighbors' current labels, averaged over classes.
    Defined in Definition 3.5; its correlation with GNN performance is demonstrated on the paper's own datasets, not on external benchmarks.
  • Dynamic compatibility matrix T_t
    purpose: Multiclass extension of dynamic homophily capturing full neighbor label distributions for future classes.
    Defined in Appendix C.1; no experiments validate it.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Understanding GNNs and Homophily in Dynamic Node Classification." pith.science (2026). https://pith.science/paper/PSDATVVC

@misc{pith2026250420421,
  author       = {Pith},
  title        = {Pith review of: Understanding GNNs and Homophily in Dynamic Node Classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PSDATVVC}},
  note         = {Machine review of arXiv:2504.20421}
}
read the original abstract

Homophily, as a measure, has been critical to increasing our understanding of graph neural networks (GNNs). However, to date this measure has only been analyzed in the context of static graphs. In our work, we explore homophily in dynamic settings. Focusing on graph convolutional networks (GCNs), we demonstrate theoretically that in dynamic settings, current GCN discriminative performance is characterized by the probability that a node's future label is the same as its neighbors' current labels. Based on this insight, we propose dynamic homophily, a new measure of homophily that applies in the dynamic setting. This new measure correlates with GNN discriminative performance and sheds light on how to potentially design more powerful GNNs for dynamic graphs. Leveraging a variety of dynamic node classification datasets, we demonstrate that popular GNNs are not robust to low dynamic homophily. Going forward, our work represents an important step towards understanding homophily and GNN performance in dynamic node classification.

Figures

Figures reproduced from arXiv: 2504.20421 by the authors.

Figure 1
Figure 1. Toy dynamic graph where the task is to predict future node labels. When [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Expected AUROC across GCN layers as a function of dynamic homophily levels. The AUROC of odd [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Mean ± standard deviations of Spearman’s rank correlation coefficient between GNN AUROC and homophily measures across graphs in the test set for a subset of datasets. For most GNN and dynamic graph combinations, dynamic homophily has a higher correlation with GNN performance compared to static homophily. graphs. Similar to the Higgs networks, all nodes appear in the train, validation, and test sets. The full details… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Mean and standard deviations of dynamic homophily, static homophily, and AUROC across all graphs [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Mean ± standard deviations of Spearman’s rank correlation coefficient between GNN AUROC and homophily measures across graphs in the test set for the remaining dynamic graph datasets. For most GNN and dynamic graph combinations, dynamic homophily has a higher correlatio…
Figure 6
Figure 6. Figure 6: Expected AUROC across GCN layers as a function of dynamic homophily levels for different µt and σ. C.3 Further Discussion of Theorem 3.4 Here, we provide additional discussion of Theorem 3.4. In [PITH_FULL_IMAGE:figures/full_fig_p025_6.png]
Figure 7
Figure 7. Figure 7: Mean and standard deviations (error bars) of dynamic homophily, static homophily, and AUROC across all graphs in the test set for the Regular, Powerlaw, and Higgs 1 dynamic graph. For all three dynamic graphs, static homophily stays relatively high across time, while d…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

51 extracted references · 35 canonical work pages

  1. [1]

    Geom-gcn: Geometric graph convolutional networks

    Hongbin Pei, Bingzhe Wei, Kevin Chen-Chuan Chang, Yu Lei, and Bo Yang. Geom-gcn: Geometric graph convolutional networks. In International Conference on Learning Representations, 2020

    work page 2020

  2. [2]

    Beyond homophily in graph neural networks: Current limitations and effective designs

    Jiong Zhu, Yujun Yan, Lingxiao Zhao, Mark Heimann, Leman Akoglu, and Danai Koutra. Beyond homophily in graph neural networks: Current limitations and effective designs. In Advances in Neural Information Processing Systems, volume 33, pages 7793--7804, 2020

    work page 2020

  3. [3]

    Unveiling the impact of local homophily on gnn fairness: In-depth analysis and new benchmarks

    Donald Loveland and Danai Koutra. Unveiling the impact of local homophily on gnn fairness: In-depth analysis and new benchmarks. In SIAM International Conference on Data Mining, 2025

    work page 2025

  4. [4]

    Is homophily a necessity for graph neural networks? In International Conference on Learning Representations, 2022

    Yao Ma, Xiaorui Liu, Neil Shah, and Jiliang Tang. Is homophily a necessity for graph neural networks? In International Conference on Learning Representations, 2022

    2022

  5. [5]

    On the impact of feature heterophily on link prediction with graph neural networks

    Jiong Zhu, Gaotang Li, Yao - An Yang, Jing Zhu, Xuehao Cui, and Danai Koutra. On the impact of feature heterophily on link prediction with graph neural networks. In Advances in Neural Information Processing Systems, 2024 a

    work page 2024

  6. [6]

    Li, Jian Tang, Guy Wolf, and Stefanie Jegelka

    Sitao Luan, Chenqing Hua, Qincheng Lu, Liheng Ma, Lirong Wu, Xinyu Wang, Minkai Xu, Xiao-Wen Chang, Doina Precup, Rex Ying, Stan Z. Li, Jian Tang, Guy Wolf, and Stefanie Jegelka. The heterophilic graph learning handbook: Benchmarks, models, theoretical analysis, applications and challenges, 2024 a . URL https://arxiv.org/abs/2407.09618

    arXiv 2024

  7. [7]

    Adaptive universal generalized pagerank graph neural network

    Eli Chien, Jianhao Peng, Pan Li, and Olgica Milenkovic. Adaptive universal generalized pagerank graph neural network. In International Conference on Learning Representations, 2021

    2021

  8. [8]

    Beyond low-frequency information in graph convolutional networks

    Deyu Bo, Xiao Wang, Chuan Shi, and Huawei Shen. Beyond low-frequency information in graph convolutional networks. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, pages 3950--3957, 2021

    work page 2021

Show all 51 references
  1. [9]

    Graph neural networks with heterophily

    Jiong Zhu, Ryan A Rossi, Anup Rao, Tung Mai, Nedim Lipka, Nesreen K Ahmed, and Danai Koutra. Graph neural networks with heterophily. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, pages 11168--11176, 2021

    2021

  2. [10]

    Two sides of the same coin: Heterophily and oversmoothing in graph convolutional neural networks

    Yujun Yan, Milad Hashemi, Kevin Swersky, Yaoqing Yang, and Danai Koutra. Two sides of the same coin: Heterophily and oversmoothing in graph convolutional neural networks. In 2022 IEEE International Conference on Data Mining (ICDM), pages 1287--1292. IEEE, 2022

    2022

  3. [11]

    Learning laplacian positional encodings for heterophilous graphs

    Michael Ito, Jiong Zhu, Dexiong Chen, Danai Koutra, and Jenna Wiens. Learning laplacian positional encodings for heterophilous graphs. In Proceedings of the International Conference on Artificial Intelligence and Statistics. PMLR , 2025

    2025

  4. [12]

    Characterizing graph datasets for node classification: Homophily-heterophily dichotomy and beyond

    Oleg Platonov, Denis Kuznedelev, Artem Babenko, and Liudmila Prokhorenkova. Characterizing graph datasets for node classification: Homophily-heterophily dichotomy and beyond. Advances in Neural Information Processing Systems, 36, 2024

    2024

  5. [13]

    Neural message passing for quantum chemistry

    Justin Gilmer, Samuel S Schoenholz, Patrick F Riley, Oriol Vinyals, and George E Dahl. Neural message passing for quantum chemistry. In International conference on machine learning, pages 1263--1272. PMLR, 2017

    2017

  6. [14]

    Kipf and Max Welling

    Thomas N. Kipf and Max Welling. Semi-supervised classification with graph convolutional networks. In International Conference on Learning Representations (ICLR), 2017

    2017

  7. [15]

    Simplifying graph convolutional networks

    Felix Wu, Amauri Souza, Tianyi Zhang, Christopher Fifty, Tao Yu, and Kilian Weinberger. Simplifying graph convolutional networks. In International conference on machine learning, pages 6861--6871. PMLR, 2019

    2019

  8. [16]

    Deeper insights into graph convolutional networks for semi-supervised learning

    Qimai Li, Zhichao Han, and Xiao-Ming Wu. Deeper insights into graph convolutional networks for semi-supervised learning. In Proceedings of the AAAI conference on artificial intelligence, volume 32, 2018

    2018

  9. [17]

    Bronstein

    Cristian Bodnar, Francesco Di Giovanni, Benjamin Paul Chamberlain, Pietro Li \`o , and Michael M. Bronstein. Neural sheaf diffusion: A topological perspective on heterophily and oversmoothing in GNN s. In Alice H. Oh, Alekh Agarwal, Danielle Belgrave, and Kyunghyun Cho, editor...

    2022

  10. [18]

    When do graph neural networks help with node classification? investigating the homophily principle on node distinguishability

    Sitao Luan, Chenqing Hua, Minkai Xu, Qincheng Lu, Jiaqi Zhu, Xiao-Wen Chang, Jie Fu, Jure Leskovec, and Doina Precup. When do graph neural networks help with node classification? investigating the homophily principle on node distinguishability. Advances in Neural Information P...

    2024

  11. [19]

    Heterophily and graph neural networks: Past, present and future

    Jiong Zhu, Yujun Yan, Mark Heimann, Lingxiao Zhao, Leman Akoglu, and Danai Koutra. Heterophily and graph neural networks: Past, present and future. IEEE Data Engineering Bulletin, 2023

    2023

  12. [20]

    Complex contagions and the weakness of long ties

    Damon Centola and Michael Macy. Complex contagions and the weakness of long ties. American journal of Sociology, 113 0 (3): 0 702--734, 2007

    2007

  13. [21]

    Complex contagions: A decade in review

    Douglas Guilbeault, Joshua Becker, and Damon Centola. Complex contagions: A decade in review. Complex spreading phenomena in social systems: Influence and contagion in real-world social networks, pages 3--25, 2018

    2018

  14. [22]

    The dynamics of viral marketing

    Jure Leskovec, Lada A Adamic, and Bernardo A Huberman. The dynamics of viral marketing. ACM Transactions on the Web (TWEB), 1 0 (1): 0 5--es, 2007

    2007

  15. [23]

    Modeling information propagation with survival theory

    Manuel Gomez-Rodriguez, Jure Leskovec, and Bernhard Sch \"o lkopf. Modeling information propagation with survival theory. In International conference on machine learning, pages 666--674. PMLR, 2013

    2013

  16. [24]

    Toward the dynamic interactome: it's about time

    Teresa M Przytycka, Mona Singh, and Donna K Slonim. Toward the dynamic interactome: it's about time. Briefings in bioinformatics, 11 0 (1): 0 15--29, 2010

    2010

  17. [25]

    Temporal networks

    Petter Holme and Jari Saram \"a ki. Temporal networks. Physics reports, 519 0 (3): 0 97--125, 2012

    2012

  18. [26]

    Patterns and dynamics of users' behavior and interaction: Network analysis of an online community

    Pietro Panzarasa, Tore Opsahl, and Kathleen M Carley. Patterns and dynamics of users' behavior and interaction: Network analysis of an online community. Journal of the American Society for Information Science and Technology, 60 0 (5): 0 911--932, 2009

    2009

  19. [27]

    Edge weight prediction in weighted signed networks

    Srijan Kumar, Francesca Spezzano, VS Subrahmanian, and Christos Faloutsos. Edge weight prediction in weighted signed networks. In Data Mining (ICDM), 2016 IEEE 16th International Conference on, pages 221--230. IEEE, 2016

    2016

  20. [28]

    Motifs in temporal networks

    Ashwin Paranjape, Austin R Benson, and Jure Leskovec. Motifs in temporal networks. In Proceedings of the tenth ACM international conference on web search and data mining, pages 601--610, 2017

    2017

  21. [29]

    A contribution to the mathematical theory of epidemics

    William Ogilvy Kermack and Anderson G McKendrick. A contribution to the mathematical theory of epidemics. Proceedings of the royal society of london. Series A, Containing papers of a mathematical and physical character, 115 0 (772): 0 700--721, 1927

    1927

  22. [30]

    Spread of epidemic disease on networks

    Mark EJ Newman. Spread of epidemic disease on networks. Physical review E, 66 0 (1): 0 016128, 2002

    2002

  23. [31]

    The anatomy of a scientific rumor

    Manlio De Domenico, Antonio Lima, Paul Mougel, and Mirco Musolesi. The anatomy of a scientific rumor. Scientific reports, 3 0 (1): 0 2980, 2013

    2013

  24. [32]

    Dppin: A biological repository of dynamic protein-protein interaction network data

    Dongqi Fu and Jingrui He. Dppin: A biological repository of dynamic protein-protein interaction network data. In 2022 IEEE International Conference on Big Data (Big Data), pages 5269--5277. IEEE, 2022

    2022

  25. [33]

    How powerful are graph neural networks? In International Conference on Learning Representations, 2019

    Keyulu Xu, Weihua Hu, Jure Leskovec, and Stefanie Jegelka. How powerful are graph neural networks? In International Conference on Learning Representations, 2019

    2019

  26. [34]

    Graph Attention Networks

    Petar Veli c kovi \' c , Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Li \` o , and Yoshua Bengio. Graph Attention Networks . International Conference on Learning Representations, 2018. accepted as poster

    2018

  27. [35]

    Hamilton, Rex Ying, and Jure Leskovec

    William L. Hamilton, Rex Ying, and Jure Leskovec. Inductive representation learning on large graphs. In Advances in Neural Information Processing Systems, 2017

    2017

  28. [36]

    Schaub, and Danai Koutra

    Donald Loveland, Jiong Zhu, Mark Heimann, Benjamin Fish, Michael T. Schaub, and Danai Koutra. On performance discrepancies across local homophily levels in graph neural networks. In Learning on Graphs Conference (LOG), volume 231 of Proceedings of Machine Learning Research, pa...

    2023

  29. [37]

    Dynamic graph convolutional networks

    Franco Manessi, Alessandro Rozza, and Mario Manzo. Dynamic graph convolutional networks. Pattern Recognition, 97: 0 107000, 2020

    2020

  30. [38]

    Learning graph dynamics using deep neural networks

    Apurva Narayan and Peter HO’N Roe. Learning graph dynamics using deep neural networks. Ifac-Papersonline, 51 0 (2): 0 433--438, 2018

    2018

  31. [39]

    Gc-lstm: Graph convolution embedded lstm for dynamic network link prediction

    Jinyin Chen, Xueke Wang, and Xuanheng Xu. Gc-lstm: Graph convolution embedded lstm for dynamic network link prediction. Applied Intelligence, pages 1--16, 2022

    2022

  32. [40]

    Structured sequence modeling with graph convolutional recurrent networks

    Youngjoo Seo, Micha \"e l Defferrard, Pierre Vandergheynst, and Xavier Bresson. Structured sequence modeling with graph convolutional recurrent networks. In Neural Information Processing: 25th International Conference, ICONIP 2018, Siem Reap, Cambodia, December 13-16, 2018, Pr...

    2018

  33. [41]

    Evolvegcn: Evolving graph convolutional networks for dynamic graphs

    Aldo Pareja, Giacomo Domeniconi, Jie Chen, Tengfei Ma, Toyotaro Suzumura, Hiroki Kanezashi, Tim Kaler, Tao Schardl, and Charles Leiserson. Evolvegcn: Evolving graph convolutional networks for dynamic graphs. In Proceedings of the AAAI conference on artificial intelligence, vol...

    2020

  34. [42]

    Finding global homophily in graph neural networks when meeting heterophily

    Xiang Li, Renyu Zhu, Yao Cheng, Caihua Shan, Siqiang Luo, Dongsheng Li, and Weining Qian. Finding global homophily in graph neural networks when meeting heterophily. In International Conference on Machine Learning, pages 13242--13256. PMLR, 2022

    2022

  35. [43]

    Ioannidis, Danai Koutra, and Christos Faloutsos

    Jing Zhu, Xiang Song, Vassilis N. Ioannidis, Danai Koutra, and Christos Faloutsos. Touchup-g: Improving feature representation through graph-centric finetuning. In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval ,...

    2024

  36. [44]

    High-dimensional statistics: A non-asymptotic viewpoint, volume 48

    Martin J Wainwright. High-dimensional statistics: A non-asymptotic viewpoint, volume 48. Cambridge university press, 2019

    2019

  37. [45]

    High-dimensional probability: An introduction with applications in data science, volume 47

    Roman Vershynin. High-dimensional probability: An introduction with applications in data science, volume 47. Cambridge university press, 2018

    2018

  38. [46]

    Generalization bounds for the area under the roc curve

    Shivani Agarwal, Thore Graepel, Ralf Herbrich, Sariel Har-Peled, Dan Roth, and Michael I Jordan. Generalization bounds for the area under the roc curve. Journal of Machine Learning Research, 6 0 (4), 2005

    2005

  39. [47]

    Large scale learning on non-homophilous graphs: New benchmarks and strong simple methods

    Derek Lim, Felix Matthew Hohne, Xiuyu Li, Sijia Linda Huang, Vaishnavi Gupta, Omkar Prasad Bhalerao, and Ser-Nam Lim. Large scale learning on non-homophilous graphs: New benchmarks and strong simple methods. In A. Beygelzimer, Y. Dauphin, P. Liang, and J. Wortman Vaughan, edit...

    2021

  40. [48]

    Predicting dynamic embedding trajectory in temporal interaction networks

    Srijan Kumar, Xikun Zhang, and Jure Leskovec. Predicting dynamic embedding trajectory in temporal interaction networks. In Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining, pages 1269--1278, 2019

    2019

  41. [49]

    Temporal graph benchmark for machine learning on temporal graphs

    Shenyang Huang, Farimah Poursafaei, Jacob Danovitch, Matthias Fey, Weihua Hu, Emanuele Rossi, Jure Leskovec, Michael Bronstein, Guillaume Rabusseau, and Reihaneh Rabbany. Temporal graph benchmark for machine learning on temporal graphs. Advances in neural information processin...

    2023

  42. [50]

    Networkx: Network analysis with python

    Aric Hagberg and Drew Conway. Networkx: Network analysis with python. URL: https://networkx. github. io, 2020

    2020

  43. [51]

    Adam: A method for stochastic optimization

    Diederik Kingma and Jimmy Ba. Adam: A method for stochastic optimization. In International Conference on Learning Representations (ICLR), San Diega, CA, USA, 2015

    2015

Pith tools

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