Pith. sign in

REVIEW 2 major objections 7 minor 70 references

Improving the Effective Receptive Field of Message-Passing Neural Networks

T0 review · 2 major / 7 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Message-passing networks lose influence with hop distance; the IM-MPNN multiscale wrapper extends their effective receptive field at linear cost.

desk verdict Solid multiscale architecture with strong experiments; the theory overclaims exponential decay and leans on an unproven grid analogy, but the core contribution survives. read the letter →

arxiv 2505.23185 v1 pith:VKV5U5BH submitted 2025-05-29 cs.LG

Shahaf E. Finder , Ron Shapira Weber , Moshe Eliasof , Oren Freifeld , Eran Treister This is my paper
classification cs.LG
keywords effectivereceptivefieldmessage-passingneuralnetworksgraphmultiscalecoarseningover-squashinglong-rangedependenciesdiffusionheterophilicnodeclassification
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 sets out to show that message-passing neural networks (MPNNs) suffer from a limited effective receptive field (ERF): the influence of a node on a distant node's output decays exponentially with hop distance, so long-range dependencies are poorly captured even when the network is deep enough in principle to reach them. It formalizes this by analyzing uniform-weight convolution on a line graph, where node contributions follow Pascal's triangle and thus a binomial distribution with exponentially small tails, and by viewing MPNN layers as a discretization of a diffusion equation whose heat-kernel solution spreads only slowly in space. To remedy this, it proposes IM-MPNN, which coarsens the graph several times, runs the same message-passing scheme on every scale, and interleaves scales with learned mixing operations, so that one step on a coarse graph reaches roughly the region that several steps reach on the original graph. The authors report that this drop-in wrapper keeps the linear O(|V|+|E|) complexity of standard MPNNs and raises accuracy on long-range benchmarks, graph-transfer tasks, and heterophilic node classification.

What carries the argument

The load-bearing mechanism is the diffusion-equation picture of message passing. The paper reads a stack of MPNN layers as a discretization of $\partial x/\partial t = -Lx$, a heat equation on the graph's latent geometry, so that the influence of a source spreads like the heat kernel of Eq. (12). The key identity is the grid-mesh relation of Eq. (13): coarsening a 2D grid from step size $h$ to $2h$ divides the discretized Laplacian $\Delta_h$ by four, so the effective diffusion constant $\kappa$ in the coarser representation multiplies by four and the Gaussian spread of Eq. (12) grows faster. That is why a single message-passing step on scale $s$ behaves like several steps on the original graph. The architecture built on this is IM-MPNN: after Graclus coarsening creates scales $G_0,\dots,G_S$, each scale runs its own MPNN with separate weights, and scale-mix layers (Eq. 20) fuse each node's features with those of its parent in the coarser graph and its children in the finer graph, so high-frequency detail stays in the fine scale while the coarse scales deliver long-range information.

What would settle it

Measure per-scale effective diffusion constants on non-grid graphs by fitting the Gaussian of Eq. (12) to empirically measured node-contribution profiles (as in Figure 1): if halving the node count does not roughly quadruple $\kappa$, the theoretical justification for the multiscale design is refuted, even though the architecture might still work empirically.

Watch

Extended reading notes

Core claim

The paper's central claim is that MPNNs have a limited effective receptive field in the same sense as CNNs: the relative contribution of a node to the output of another node decays exponentially with the graph distance between them. On a linear graph, $\ell$ uniform convolution steps yield binomial coefficients, and Hoeffding's inequality bounds the total contribution of a block of far-away nodes by $\exp(-2(1/2-k/\ell)^2 \ell)$; in the diffusion picture, the point-source solution $x(p,t)=(4\kappa\pi t)^{-d/2}\exp(-\lVert p-p_0\rVert^2/(4\kappa t))$ spreads only as a Gaussian whose variance grows linearly with the number of layers. Because this spread is slow, stacking layers is an inefficient route to long-range reach. IM-MPNN counters this by coarsening the graph $S$ times with Graclus pooling, running the chosen MPNN backbone on each of the $S+1$ scales, and interleaving the scales with learned mixing so each node receives features from its parent and children nodes. With width reduced to match the parameter budget of the baseline, IM-MPNN reports up to 41% relative improvement on PascalVOC-SP and COCO-SP, near-perfect transfer of a label across 37 hops on CliquePath where a standard GCN fails near 7 hops, and accuracy gains of 10 to 16.5 points on City-Networks, while keeping the total cost linear in graph size.

Load-bearing premise

The claim that coarser scales widen the receptive field depends on treating a coarsened graph as a factor-two-coarser discretization of the same continuous Laplacian, so that the diffusion constant $\kappa$ quadruples — an analogy proven for regular grid-like meshes but not established for arbitrary graphs.

Editorial extensions

If this is right

  • One message-passing step on the $s$-th coarsened scale reaches about $2^s$ hops of the original graph, so far fewer layers are needed to move information across a large graph.
  • Any linear-complexity MPNN backbone can be wrapped by IM-MPNN while keeping total complexity linear, because the work across all scales sums to $O(2(|V|+|E|))$.
  • Deepening a plain MPNN is a comparatively inefficient way to gain receptive field, since long-range contributions decay exponentially; multiscale processing reuses the same parameters more effectively.
  • The reported gains — up to 41% relative on PascalVOC-SP and COCO-SP, 10–16.5 accuracy points on City-Networks, and 100% label-transfer at 37 hops where GCN fails at 7 — indicate that many long-range failures of MPNNs stem from the limited ERF rather than from insufficient expressivity.
  • The number of scales is a tunable hyperparameter, and adding scales returns diminishing gains once the coarsest graph is reached or when the graph is highly homophilic.

Reading between the lines

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

  • Testable extension (ours): compare IM-MPNN against a single-scale MPNN given the same parameter budget and against one augmented with distance or random-walk encodings; if the multiscale wrapper still wins clearly, the ERF mechanism is doing real work rather than merely adding capacity.
  • Editorial reading: the learned scale-mix weights $W_{l2h}$ and $W_{h2l}$ act as per-scale diffusion constants; monitoring their magnitudes during training could reveal whether the model genuinely channels long-range information through the coarse scales.
  • The paper uses Graclus pooling, but its own coarsening-agnostic design implies that any pairing scheme that roughly halves graph distances — spectral or learned pooling, for instance — should extend the ERF similarly, making the gains track the distance-reduction factor of the pooling.
  • A natural next step the authors leave implicit is composing IM-MPNN with graph rewiring or graph-transformer layers: the coarse scales supply a cheap global channel that could complement attention rather than compete with it.
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

2 major / 7 minor

Summary. The paper studies the effective receptive field (ERF) of message-passing neural networks (MPNNs). It argues that the contribution of a node to a distant target decays rapidly, characterizes this through a linear-graph binomial analysis and a continuous diffusion analogy, and proposes IM-MPNN, an architecture that runs message passing on several coarsened graphs and interleaves information between scales. The method is evaluated on the Long-Range Graph Benchmark, a graph transfer task, City-Networks, and heterophilic node-classification benchmarks, with several MPNN backbones, controlled parameter budgets, and multiple random seeds. The results show consistent improvements over the corresponding single-scale baselines, especially on long-range tasks, while maintaining linear complexity.

Significance. If the empirical results hold, IM-MPNN is a simple, general, and practical enhancement for MPNNs on long-range tasks: it requires no change to the backbone aside from adding coarsening and scale-mixing, it respects parameter budgets, and the reported gains on PascalVOC-SP, COCO-SP, City-Networks, and graph transfer are substantial. The paper ships code, reports standard deviations over multiple seeds, and controls for parameter count, which strengthens the empirical contribution. The theoretical part is weaker: the claimed 'exponential' decay is actually Gaussian in distance for the linear-graph model, and the factor-four coarsening argument in Section 4.1 is only proven for regular grids, not for the unweighted Graclus coarsening used in practice. These issues affect the paper's explanatory narrative but do not invalidate the central empirical claim.

major comments (2)
  1. [Section 3.1, Eqs. (5)-(8)] The claim that the relative contribution of nodes decreases 'exponentially with their distance' is not supported by the binomial analysis. For a fixed depth ℓ, the binomial coefficient at displacement d from the center behaves like (2/(πℓ))^{1/2} exp(−d^2/(2ℓ)), which is a Gaussian, not an exponential, decay in distance. Hoeffding's inequality in Eq. (8) bounds a cumulative tail for k = αℓ, so it decays exponentially in ℓ, not per-node exponentially in d. The qualitative conclusion that the ERF is much smaller than the theoretical receptive field remains valid, since the standard deviation grows only as √ℓ, but the abstract, Section 3, and conclusion should be corrected from 'exponential' to 'Gaussian/squared-exponential' decay.
  2. [Section 4.1, Eq. (13)] The argument that factor-two coarsening multiplies the diffusion constant κ by four rests on the 2D four-neighbor grid kernel with uniform step h, and is not established for the unweighted Graclus coarsening used in the implementation. Quotient graphs have variable degrees and no global coordinate step, so the graph Laplacians at different scales are not simply scaled versions of one continuous operator with a global κ. Since this factor-four scaling is the paper's stated mechanism for why coarsening expands the ERF, the authors should either prove the scaling for the actual coarsening procedure or explicitly downgrade it to heuristic motivation and rely instead on the direct observation that coarsening shortens graph distances. The empirical results do not depend on this argument, but the theoretical explanation in Section 4.1 does.
minor comments (7)
  1. [Eq. (15)] The condition for E′ should read ∃ vi ∈ v′q, vj ∈ v′p, (vi, vj) ∈ E; as written, both vertices are drawn from the same component v′q, which contradicts the surrounding text.
  2. [Section 4.1] The phrase 'which we present in Section 4' is self-referential; it should be 'Section 4.2'.
  3. [Eq. (20)] The boundary cases s = 0 and s = S are not specified; please state explicitly that the lower-to-higher or higher-to-lower terms are omitted at the finest and coarsest scales.
  4. [Section 5.1] The text says 'up to 41% relative improvement on both PascalVOC-SP and COCO-SP', but COCO-SP improves from 0.1338 to 0.1960, which is about 46.5% relative; please make the reported number consistent.
  5. [Figure 1] The procedure used to measure the node contributions is not described; please add a short description (or an appendix entry) so that the ERF visualization is reproducible.
  6. [Typos] There are two typos: 'V onessen' in Section 2 before the reference should be 'Vonessen', and 'ORGB Arxiv' in Table 10 should be 'OGBN Arxiv'.
  7. [Appendix C.4] The selected number of scales for each dataset and seed is not reported; please include this hyperparameter setting in the appendix.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the ERF analysis is derived from explicit recurrences and standard heat-kernel results, and the architecture is validated against external benchmarks rather than against the paper's own fitted quantities.

full rationale

The derivation chain is self-contained. Section 3.1 builds the exponential-decay claim from explicit message-passing recurrences (Eqs. 1-5), normalizes to a binomial distribution, and bounds the tail via Hoeffding (Eqs. 6-8); no target result is assumed. Section 3.2 invokes the standard heat-kernel solution (Eq. 12) under stated assumptions about graph-Laplacian discretization, and the two stated insights follow directly from the Gaussian form. Section 4.1's factor-four scaling is derived by the grid Laplacian identity (Eq. 13) rather than by fitting; the extension to Graclus-coarsened arbitrary graphs is an unproven analogy, which is a motivation or correctness gap, not a circular reduction. The empirical claims on LRGB, graph transfer, City-Networks, and heterophilic node classification are measured against external baselines, not against quantities derived from the paper's own model. The sole self-citation (Finder et al., 2024) is cited only as inspiration for multiscale CNN ERF augmentation and is not load-bearing for any theorem or architecture claim. Appendix B explicitly acknowledges that the number of scales is a validation-tuned hyperparameter, which is standard tuning rather than a fitted parameter renamed as a prediction. No self-definitional, fitted-input-as-prediction, imported-uniqueness, or ansatz-by-citation pattern is present.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The main free parameter is the number of scales S, which is tuned per dataset. The theoretical arguments rely on the diffusion analogy and the scale-equivalence assumption for coarsening, neither of which is rigorously established for general graphs. No new physical or conceptual entities are introduced.

free parameters (1)
  • Number of scales S = 1-4 (per dataset)
    Appendix B states the number of scales is a hyperparameter requiring empirical tuning; experiments sweep S=1..4 (and 2,3,4,8 for heterophilic) and select per dataset.
assumptions (4)
  • domain assumption The binomial coefficient counting in Eq. (5) for a path graph with uniform weights models real MPNN behavior.
    Section 3.1 analyzes an infinitely-long linear graph with no self-loops and uniform weights; real MPNNs use non-uniform learned weights, self-loops, and nonlinearities, so the decay rates differ.
  • domain assumption Graph Laplacians at multiple coarsening scales are discretizations of the same continuous Laplacian, with the effective diffusion constant kappa scaling by 4 per halving.
    Section 4.1 derives the ERF expansion argument from a 2D grid mesh analogy (Eq. 13); no proof is given for arbitrary graphs.
  • domain assumption The heat-kernel solution in Eq. (12) applies to discrete graph diffusion.
    Section 3.2 assumes the graph Laplacian discretizes the continuous Laplace operator under some geometry with node coordinates p, which is an approximation for general graphs.
  • standard math Hoeffding's inequality provides a valid tail bound for the binomial distribution (used in Eq. 8).
    The paper applies Hoeffding's inequality to the normalized binomial coefficients; this is standard probability theory.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Improving the Effective Receptive Field of Message-Passing Neural Networks." pith.science (2026). https://pith.science/paper/VKV5U5BH

@misc{pith2026250523185,
  author       = {Pith},
  title        = {Pith review of: Improving the Effective Receptive Field of Message-Passing Neural Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VKV5U5BH}},
  note         = {Machine review of arXiv:2505.23185}
}
read the original abstract

Message-Passing Neural Networks (MPNNs) have become a cornerstone for processing and analyzing graph-structured data. However, their effectiveness is often hindered by phenomena such as over-squashing, where long-range dependencies or interactions are inadequately captured and expressed in the MPNN output. This limitation mirrors the challenges of the Effective Receptive Field (ERF) in Convolutional Neural Networks (CNNs), where the theoretical receptive field is underutilized in practice. In this work, we show and theoretically explain the limited ERF problem in MPNNs. Furthermore, inspired by recent advances in ERF augmentation for CNNs, we propose an Interleaved Multiscale Message-Passing Neural Networks (IM-MPNN) architecture to address these problems in MPNNs. Our method incorporates a hierarchical coarsening of the graph, enabling message-passing across multiscale representations and facilitating long-range interactions without excessive depth or parameterization. Through extensive evaluations on benchmarks such as the Long-Range Graph Benchmark (LRGB), we demonstrate substantial improvements over baseline MPNNs in capturing long-range dependencies while maintaining computational efficiency.

Figures

Figures reproduced from arXiv: 2505.23185 by the authors.

Figure 1
Figure 1. Measuring the contribution of each node to the [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. IM-MPNN architecture for scales=3. The input is first passed through an encoding stage (PE, SE, Graph features, [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. An infinitely-long linear graph. efficient and scalable message-passing, and Eliasof et al. (2023a) develop a wavelet based multiscale approach for the compression of node features in GNNs. Lastly, in the con￾text of beyond MPNNs, Luo (2023) and Zhang et al. (2022) introduced a hierarchical encoding mechanism for graph transformers, effectively capturing multi-scale graph struc￾ture using hierarchical distances to i… view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: The spread of a point source in time according to [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: Coarsening of a graph according to a given pairing [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Scale-mix. Each node receives information from [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: Graph transfer results over three graph types (ring, crossed-ring, and cliquepath). The network depth (i.e. number [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Line, ring, and crossed-ring graphs where the distance between source and target nodes is equal to 5. Nodes [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

70 extracted references · 55 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

  2. [2]

    Mixhop: Higher-order graph convolutional architectures via sparsified neighborhood mixing

    Abu-El-Haija, S., Perozzi, B., Kapoor, A., Alipourfard, N., Lerman, K., Harutyunyan, H., Ver Steeg, G., and Galstyan, A. Mixhop: Higher-order graph convolutional architectures via sparsified neighborhood mixing. In international conference on machine learning, 2019

    work page 2019

  3. [3]

    Slic superpixels, 2010

    Achanta, R., Shaji, A., Smith, K., Lucchi, A., Fua, P., and S \"u sstrunk, S. Slic superpixels, 2010

    work page 2010

  4. [4]

    and Yahav, E

    Alon, U. and Yahav, E. On the bottleneck of graph neural networks and its practical implications. In International Conference on Learning Representations, 2021

    work page 2021

  5. [5]

    and Hashemi, F

    Behrouz, A. and Hashemi, F. Graph Mamba: Towards Learning on Graphs with State Space Models . arXiv preprint arXiv:2402.08678, 2024

    arXiv 2024

  6. [6]

    and Niyogi, P

    Belkin, M. and Niyogi, P. Laplacian eigenmaps and spectral techniques for embedding and clustering. Advances in neural information processing systems, 14, 2001

    work page 2001

  7. [7]

    Beyond low-frequency information in graph convolutional networks

    Bo, D., Wang, X., Shi, C., and Shen, H. Beyond low-frequency information in graph convolutional networks. Proceedings of the AAAI Conference on Artificial Intelligence, 35, 2021. doi:10.1609/aaai.v35i5.16514

    doi:10.1609/aaai.v35i5.16514 2021

  8. [8]

    and Laurent, T

    Bresson, X. and Laurent, T. Residual gated graph convnets. arXiv preprint arXiv:1711.07553, 2017

    arXiv 2017

Show all 70 references
  1. [9]

    B., Lodi, A., Morris, C., and Veli c kovi \'c , P

    Cappart, Q., Ch \'e telat, D., Khalil, E. B., Lodi, A., Morris, C., and Veli c kovi \'c , P. Combinatorial optimization and reasoning with graph neural networks. Journal of Machine Learning Research, 24 0 (130): 0 1--61, 2023

    2023

  2. [10]

    Beltrami flow and neural diffusion on graphs

    Chamberlain, B., Rowbottom, J., Eynard, D., Di Giovanni, F., Dong, X., and Bronstein, M. Beltrami flow and neural diffusion on graphs. Advances in Neural Information Processing Systems, 34: 0 1594--1609, 2021 a

    2021

  3. [11]

    P., Rowbottom, J., Gorinova, M., Webb, S., Rossi, E., and Bronstein, M

    Chamberlain, B. P., Rowbottom, J., Gorinova, M., Webb, S., Rossi, E., and Bronstein, M. M. GRAND : Graph neural diffusion. In International Conference on Machine Learning, 2021 b

    2021

  4. [12]

    Improving message-passing gnns by asynchronous aggregation

    Chen, J., Liao, T., Chen, C., and Zheng, Z. Improving message-passing gnns by asynchronous aggregation. In Proceedings of the 33rd ACM International Conference on Information and Knowledge Management, pp.\ 228--238, 2024

    2024

  5. [13]

    Adaptive universal generalized pagerank graph neural network

    Chien, E., Peng, J., Li, P., and Milenkovic, O. Adaptive universal generalized pagerank graph neural network. In International Conference on Learning Representations, 2021

    2021

  6. [14]

    Gread: Graph neural reaction-diffusion networks

    Choi, J., Hong, S., Park, N., and Cho, S.-B. Gread: Graph neural reaction-diffusion networks. In International Conference on Machine Learning, pp.\ 5722--5747. PMLR, 2023

    2023

  7. [15]

    S., Guan, Y., and Kulis, B

    Dhillon, I. S., Guan, Y., and Kulis, B. Weighted graph cuts without eigenvectors a multilevel approach. IEEE transactions on pattern analysis and machine intelligence, 29 0 (11): 0 1944--1957, 2007

    1944

  8. [16]

    Di Giovanni, F., Giusti, L., Barbero, F., Luise, G., Lio, P., and Bronstein, M. M. On over-squashing in message passing neural networks: The impact of width, depth, and topology. In International Conference on Machine Learning, pp.\ 7865--7885. PMLR, 2023

    2023

  9. [17]

    Scaling up your kernels to 31x31: Revisiting large kernel design in cnns

    Ding, X., Zhang, X., Han, J., and Ding, G. Scaling up your kernels to 31x31: Revisiting large kernel design in cnns. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp.\ 11963--11975, 2022

    2022

  10. [18]

    Gbk-gnn: Gated bi-kernel graph neural networks for modeling both homophily and heterophily

    Du, L., Shi, X., Fu, Q., Ma, X., Liu, H., Han, S., and Zhang, D. Gbk-gnn: Gated bi-kernel graph neural networks for modeling both homophily and heterophily. In Proceedings of the ACM Web Conference 2022, pp.\ 1550--1558, 2022

    2022

  11. [19]

    Dwivedi, V. P. and Bresson, X. A generalization of transformer networks to graphs. AAAI Workshop on Deep Learning on Graphs: Methods and Applications, 2021

    2021

  12. [20]

    P., Ramp \'a s ek, L., Galkin, M., Parviz, A., Wolf, G., Luu, A

    Dwivedi, V. P., Ramp \'a s ek, L., Galkin, M., Parviz, A., Wolf, G., Luu, A. T., and Beaini, D. Long range graph benchmark. Advances in Neural Information Processing Systems, 35: 0 22326--22340, 2022

    2022

  13. [21]

    J., and Treister, E

    Eliasof, M., Bodner, B. J., and Treister, E. Haar wavelet feature compression for quantized graph convolutional networks. IEEE Transactions on Neural Networks and Learning Systems, 2023 a

    2023

  14. [22]

    Improving graph neural networks with learnable propagation operators

    Eliasof, M., Ruthotto, L., and Treister, E. Improving graph neural networks with learnable propagation operators. In International Conference on Machine Learning, pp.\ 9224--9245. PMLR, 2023 b

    2023

  15. [23]

    K., Winn, J., and Zisserman, A

    Everingham, M., Van Gool, L., Williams, C. K., Winn, J., and Zisserman, A. The pascal visual object classes (voc) challenge. International journal of computer vision, 88: 0 303--338, 2010

    2010

  16. [24]

    Graph neural networks for social recommendation

    Fan, W., Ma, Y., Li, Q., He, Y., Zhao, E., Tang, J., and Yin, D. Graph neural networks for social recommendation. In The world wide web conference, pp.\ 417--426, 2019

    2019

  17. [25]

    E., Amoyal, R., Treister, E., and Freifeld, O

    Finder, S. E., Amoyal, R., Treister, E., and Freifeld, O. Wavelet convolutions for large receptive fields. In European Conference on Computer Vision, pp.\ 363--380. Springer, 2024

    2024

  18. [26]

    M., and Ceylan, I

    Finkelshtein, B., Huang, X., Bronstein, M. M., and Ceylan, I. I. Cooperative graph neural networks. In International Conference on Machine Learning, pp.\ 13633--13659. PMLR, 2024

    2024

  19. [27]

    and Ji, S

    Gao, H. and Ji, S. Graph u-nets. In international conference on machine learning, pp.\ 2083--2092. PMLR, 2019

    2019

  20. [28]

    Diffusion improves graph learning

    Gasteiger, J., Wei enberger, S., and G \"u nnemann, S. Diffusion improves graph learning. Advances in neural information processing systems, 32, 2019

    2019

  21. [29]

    Tackling graph oversquashing by global and local non-dissipativity

    Gravina, A., Eliasof, M., Gallicchio, C., Bacciu, D., and Sch \"o nlieb, C.-B. Tackling graph oversquashing by global and local non-dissipativity. arXiv preprint arXiv:2405.01009, 2024

    2024 arXiv

  22. [30]

    M., and Di Giovanni, F

    Gutteridge, B., Dong, X., Bronstein, M. M., and Di Giovanni, F. Drew: Dynamically rewired message passing with delay. In International Conference on Machine Learning, 2023

    2023

  23. [31]

    Inductive representation learning on large graphs

    Hamilton, W., Ying, Z., and Leskovec, J. Inductive representation learning on large graphs. Advances in neural information processing systems, 30, 2017

    2017

  24. [32]

    From continuous dynamics to graph neural networks: Neural diffusion and beyond

    Han, A., Shi, D., Lin, L., and Gao, J. From continuous dynamics to graph neural networks: Neural diffusion and beyond. Transactions on Machine Learning Research, 2024. ISSN 2835-8856

    2024

  25. [33]

    Strategies for pre-training graph neural networks

    Hu, W., Liu, B., Gomes, J., Zitnik, M., Liang, P., Pande, V., and Leskovec, J. Strategies for pre-training graph neural networks. In International Conference on Learning Representations, 2020

    2020

  26. [34]

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

    2016 arXiv

  27. [35]

    B., and Goldstein, T

    Kong, K., Chen, J., Kirchenbauer, J., Ni, R., Bruss, C. B., and Goldstein, T. GOAT : A global transformer on large-scale graphs. In Proceedings of the 40th International Conference on Machine Learning, 2023

    2023

  28. [36]

    Rethinking graph transformers with spectral attention

    Kreuzer, D., Beaini, D., Hamilton, W., L \'e tourneau, V., and Tossou, P. Rethinking graph transformers with spectral attention. Advances in Neural Information Processing Systems, 34: 0 21618--21629, 2021

    2021

  29. [37]

    Finding global homophily in graph neural networks when meeting heterophily

    Li, X., Zhu, R., Cheng, Y., Shan, C., Luo, S., Li, D., and Qian, W. Finding global homophily in graph neural networks when meeting heterophily. In Proceedings of the 39th International Conference on Machine Learning, 2022

    2022

  30. [38]

    Liang, H., Borde, H. S. d. O., Sripathmanathan, B., Bronstein, M., and Dong, X. Towards quantifying long-range interactions in graph machine learning: a large graph dataset and a measurement. arXiv preprint arXiv:2503.09008, 2025

    2025

  31. [39]

    Toloker graph: Interaction of crowd annotators, 2023

    Likhobaba, D., Pavlichenko, N., and Ustalov, D. Toloker graph: Interaction of crowd annotators, 2023

    2023

  32. [40]

    Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Doll \'a r, P., and Zitnick, C. L. Microsoft coco: Common objects in context. In Computer Vision--ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part V 13, pp....

    2014

  33. [41]

    Geniepath: Graph neural networks with adaptive receptive paths

    Liu, Z., Chen, C., Li, L., Zhou, J., Li, X., Song, L., and Qi, Y. Geniepath: Graph neural networks with adaptive receptive paths. In Proceedings of the AAAI Conference on Artificial Intelligence, 2019

    2019

  34. [42]

    Understanding the effective receptive field in deep convolutional neural networks

    Luo, W., Li, Y., Urtasun, R., and Zemel, R. Understanding the effective receptive field in deep convolutional neural networks. Advances in neural information processing systems, 29, 2016

    2016

  35. [43]

    Transformers for capturing multi-level graph structure using hierarchical distances

    Luo, Y. Transformers for capturing multi-level graph structure using hierarchical distances. CoRR, 2023

    2023

  36. [44]

    Improving graph neural networks with structural adaptive receptive fields

    Ma, X., Wang, J., Chen, H., and Song, G. Improving graph neural networks with structural adaptive receptive fields. In Proceedings of the Web Conference 2021, pp.\ 2438--2447, 2021

    2021

  37. [45]

    Learning discrete adaptive receptive fields for graph convolutional networks

    Ma, X., Li, Z., Song, G., and Shi, C. Learning discrete adaptive receptive fields for graph convolutional networks. Science China Information Sciences, 66 0 (12): 0 222101, 2023

    2023

  38. [46]

    QDC : Quantum diffusion convolution kernels on graphs

    Markovich, T. QDC : Quantum diffusion convolution kernels on graphs. Transactions on Machine Learning Research, 2024

    2024

  39. [47]

    A fractional graph laplacian approach to oversmoothing

    Maskey, S., Paolino, R., Bacho, A., and Kutyniok, G. A fractional graph laplacian approach to oversmoothing. Advances in Neural Information Processing Systems, 36: 0 13022--13063, 2023

    2023

  40. [48]

    K., Liu, X., and Murata, T

    Maurya, S. K., Liu, X., and Murata, T. Simplifying approach to node classification in graph neural networks. Journal of Computational Science, 62: 0 101695, 2022

    2022

  41. [49]

    Attending to graph transformers

    M \"u ller, L., Galkin, M., Morris, C., and Ramp \'a s ek, L. Attending to graph transformers. Transactions on Machine Learning Research, 2024. ISSN 2835-8856

    2024

  42. [50]

    and Duta, I

    Nicolicioiu, A. and Duta, I. Effective receptive field of graph neural networks, 2019

    2019

  43. [51]

    and Maehara, T

    Nt, H. and Maehara, T. Revisiting graph neural networks: All we have is low-pass filters. arXiv preprint arXiv:1905.09550, 2019

    1905 arXiv

  44. [52]

    Pattle, R. E. Diffusion from an instantaneous point source with a concentration-dependent coefficient. The Quarterly Journal of Mechanics and Applied Mathematics, 12 0 (4): 0 407--409, 1959

    1959

  45. [53]

    A critical look at the evaluation of GNN s under heterophily: Are we really making progress? In The Eleventh International Conference on Learning Representations, 2023

    Platonov, O., Kuznedelev, D., Diskin, M., Babenko, A., and Prokhorenkova, L. A critical look at the evaluation of GNN s under heterophily: Are we really making progress? In The Eleventh International Conference on Learning Representations, 2023

    2023

  46. [54]

    P., Luu, A

    Ramp \'a s ek, L., Galkin, M., Dwivedi, V. P., Luu, A. T., Wolf, G., and Beaini, D. Recipe for a general, powerful, scalable graph transformer. Advances in Neural Information Processing Systems, 35: 0 14501--14515, 2022

    2022

  47. [55]

    Masked label prediction: Unified message passing model for semi-supervised classification

    Shi, Y., Huang, Z., Feng, S., Zhong, H., Wang, W., and Sun, Y. Masked label prediction: Unified message passing model for semi-supervised classification. In Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence, pp.\ 1548--1554. International J...

    2021

  48. [56]

    J., and Sinop, A

    Shirzad, H., Velingker, A., Venkatachalam, B., Sutherland, D. J., and Sinop, A. K. Exphormer: Sparse transformers for graphs. In International Conference on Machine Learning, pp.\ 31613--31632. PMLR, 2023

    2023

  49. [57]

    Graph neural networks in particle physics

    Shlomi, J., Battaglia, P., and Vlimant, J.-R. Graph neural networks in particle physics. Machine Learning: Science and Technology, 2 0 (2): 0 021001, 2020

    2020

  50. [58]

    Where did the gap go? reassessing the long-range graph benchmark

    T \"o nshoff, J., Ritzert, M., Rosenbluth, E., and Grohe, M. Where did the gap go? reassessing the long-range graph benchmark. arXiv preprint arXiv:2309.00367, 2023

    2023 arXiv

  51. [59]

    P., Dong, X., and Bronstein, M

    Topping, J., Di Giovanni, F., Chamberlain, B. P., Dong, X., and Bronstein, M. M. Understanding over-squashing and bottlenecks on graphs via curvature. In International Conference on Learning Representations, 2022

    2022

  52. [60]

    Capturing graphs with hypo-elliptic diffusions

    Toth, C., Lee, D., Hacker, C., and Oberhauser, H. Capturing graphs with hypo-elliptic diffusions. Advances in Neural Information Processing Systems, 35: 0 38803--38817, 2022

    2022

  53. [61]

    Graph attention networks

    Veli c kovi \'c , P., Cucurull, G., Casanova, A., Romero, A., Li \`o , P., and Bengio, Y. Graph attention networks. In International Conference on Learning Representations, 2018

    2018

  54. [62]

    Next level message-passing with hierarchical support graphs

    Vonessen, C., Gr \"o tschla, F., and Wattenhofer, R. Next level message-passing with hierarchical support graphs. arXiv preprint arXiv:2406.15852, 2024

    2024 arXiv

  55. [63]

    and Zhang, M

    Wang, X. and Zhang, M. How powerful are spectral graph neural networks. In International conference on machine learning, pp.\ 23341--23362. PMLR, 2022

    2022

  56. [64]

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

    Xu, K., Hu, W., Leskovec, J., and Jegelka, S. How powerful are graph neural networks? In International Conference on Learning Representations, 2019

    2019

  57. [65]

    Sebot: Structural entropy guided multi-view contrastive learning for social bot detection

    Yang, Y., Wu, Q., He, B., Peng, H., Yang, R., Hao, Z., and Liao, Y. Sebot: Structural entropy guided multi-view contrastive learning for social bot detection. In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pp.\ 3841--3852, 2024

    2024

  58. [66]

    Hierarchical graph representation learning with differentiable pooling

    Ying, Z., You, J., Morris, C., Ren, X., Hamilton, W., and Leskovec, J. Hierarchical graph representation learning with differentiable pooling. Advances in neural information processing systems, 31, 2018

    2018

  59. [67]

    Hierarchical graph transformer with adaptive node sampling

    Zhang, Z., Liu, Q., Hu, Q., and Lee, C.-K. Hierarchical graph transformer with adaptive node sampling. Advances in Neural Information Processing Systems, 35: 0 21171--21183, 2022

    2022

  60. [68]

    Hierarchical message-passing graph neural networks

    Zhong, Z., Li, C.-T., and Pang, J. Hierarchical message-passing graph neural networks. Data Mining and Knowledge Discovery, 37 0 (1): 0 381--408, 2023

    2023

  61. [69]

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

    Zhu, J., Yan, Y., Zhao, L., Heimann, M., Akoglu, L., and Koutra, D. Beyond homophily in graph neural networks: Current limitations and effective designs. In Advances in Neural Information Processing Systems, 2020

    2020

  62. [70]

    A., Rao, A., Mai, T., Lipka, N., Ahmed, N

    Zhu, J., Rossi, R. A., Rao, A., Mai, T., Lipka, N., Ahmed, N. K., and Koutra, D. Graph neural networks with heterophily. Proceedings of the AAAI Conference on Artificial Intelligence, 2021. doi:10.1609/aaai.v35i12.17332

    2021 doi

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.