Pith. sign in

REVIEW 4 major objections 6 minor 49 references

GraphMoRE: Mitigating Topological Heterogeneity via Mixture of Riemannian Experts

T0 review · 4 major / 6 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 By routing every node through a learned mixture of hyperbolic, Euclidean, and spherical spaces, GraphMoRE claims to embed graphs into a point-dependent curvature manifold and to beat all compared baselines on link prediction.

desk verdict Real novelty and consistent empirical wins, but the 'heterogeneous manifold' is a label, not a constructed geometry; fixable in revision. read the letter →

arxiv 2412.11085 v1 pith:2F2PTLOV submitted 2024-12-15 cs.LG cs.AI

Zihao Guo , Qingyun Sun , Haonan Yuan , Xingcheng Fu , Min Zhou , Yisen Gao , Jianxin Li This is my paper · ORCID
classification cs.LGcs.AI
keywords graphrepresentationlearningtopologicalheterogeneityRiemannianmanifoldmixtureofexpertsκ-stereographicmodellinkpredictionnodeclassificationneuralnetworks
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

GraphMoRE is an attempt to solve the graph-embedding problem caused by topological heterogeneity by giving each node its own mixture of constant-curvature spaces instead of embedding the whole graph in one space. The authors build a Mixture-of-Experts model in which every expert is a graph neural network that operates in the κ-stereographic model with a fixed curvature, spanning hyperbolic, Euclidean, and spherical geometries. A topology-aware gating network estimates the local geometry around each node from multi-resolution sampled subgraphs and produces expert weights for that node; the node's embedding is the gated fusion of all expert outputs, and an alignment rule defines pairwise distances between nodes living in different mixtures. The model is trained jointly on the downstream task and a distortion-minimization term, and the paper reports that it beats nine baselines on link prediction on all five real datasets, wins most node-classification settings, and achieves the lowest average embedding distortion on every dataset. The broader interest is that a single heterogeneous geometry of this kind could let one graph foundation model handle many graph types uniformly.

What carries the argument

The central object is the personalized mixed-curvature embedding space, produced by a gated mixture of Riemannian experts and measured by an aligned weighted distance. Each expert is a graph neural network built on the κ-stereographic model, which is the named unifying object: one smooth manifold family $M^d_\kappa$ that becomes the Poincaré ball for $\kappa<0$, Euclidean space for $\kappa=0$, and the stereographic sphere for $\kappa>0$, so exponent and logarithm maps are available across all experts. The topology-aware gating mechanism, composed of multi-resolution local-topology encoding plus a distortion-guided gating network, carries the routing; the distortion loss $L_D = \frac{1}{|V|^2} \sum_{i,j} \left| \left(\frac{d(i,j)}{g(i,j)}\right)^2 - 1 \right|$ steers each node toward the expert mixtures that preserve its local geometry. The alignment strategy in Eq. (9)-(10) is the mechanism that makes distances between nodes in different mixtures comparable.

What would settle it

Compute the aligned distance from Eq. (10) for every triple in a small graph and check the metric axioms: if any triple violates the triangle inequality or has $d(u,u)\neq 0$ for embedded nodes, the mixed space is not a Riemannian metric space and the heterogeneous-manifold claim fails. A second decisive test is to ablate Eq. (9) by fixing $W_{(u,v)}$ to the uniform weight vector while retraining; if AUC on Cora or Citeseer does not drop, the alignment strategy is not load-bearing for the reported gains.

Watch

Extended reading notes

Core claim

The central claim is that local topological heterogeneity is better served by personalized mixed-curvature spaces than by any globally uniform space, including a product manifold. GraphMoRE realizes this claim by treating each curvature sign and curving degree as an expert: the experts are Riemannian GNNs with fixed curvatures $\{-3,-1,0,1,3\}$, the gating network reads sampled subgraphs at multiple radii and outputs per-node weights, and the node embedding is the fusion $Z_v = \mathrm{concat}_{\varepsilon\in\mathcal{E}}(W^\varepsilon_v \otimes_\varepsilon Z^\varepsilon_v)$. To compare nodes in different mixtures, the paper proposes $W_{(u,v)} = \mathrm{Softmax}(W_u \cdot W_v)$ and $d^2(u,v) = \sum_{\varepsilon\in\mathcal{E}} W^\varepsilon_{(u,v)} d^2_\varepsilon(Z^\varepsilon_u, Z^\varepsilon_v)$. The authors argue that this construction embeds the graph into a heterogeneous manifold whose curvature varies from point to point, and they report that it outperforms all nine baselines on link prediction on Cora, Citeseer, Airport, PubMed, and Photo, with the lowest average embedding distortion on each.

Load-bearing premise

Everything rests on treating the weighted combination of expert embeddings plus the alignment rule as a real geometric space; no metric, distance-triangle law, or coordinate geometry is ever proved for that space, so if those operations do not form a genuine geometry, the method's deeper claim reduces to a useful weighted-distance trick.

Editorial extensions

If this is right

  • If the claims hold, graph embeddings no longer need a globally fixed curvature: every node can receive its own mixture, and the model reports higher link prediction AUC than all nine baselines on Cora, Citeseer, Airport, PubMed, and Photo.
  • On Cora, Citeseer, and PubMed the reported AUC gains over the runner-up are 2.01, 1.94, and 1.92 percentage points, so the advantage is not marginal on citation networks.
  • The ablation study ties every major component to the gains: removing the distortion loss, gating, expert diversity, or alignment each lowers AUC and F1, with loss of expert diversity producing the largest drop (Cora AUC from 97.91 to 94.34).
  • Since GraphMoRE keeps its accuracy on larger synthetic graphs while baselines degrade, per-node routing appears to decouple performance from graph scale.
  • The same architecture wins most node-classification settings with GCN, GAT, and SAGE backbones, indicating the method is not tied to one classifier.

Reading between the lines

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

  • Editorial inference: the method's geometric core is testable in a formal follow-up: one could define the tangent space and geodesic of the mixed space from the gated expert maps and verify that the claim of point-varying curvature is a genuine Riemannian structure rather than a weighted-distance heuristic.
  • Editorial inference: feeding explicit discrete curvature estimators, such as graph sectional curvature, into the gating network would make the routing interpretable and reveal whether the learned weights track geometric measures; this is not tested in the paper.
  • Editorial inference: zero-shot routing to unseen graphs would determine whether GraphMoRE's per-node curvature mixture generalizes across datasets, a property graph foundation models would need.
  • Editorial inference: the alignment rule Eq. (9) can be read as learned metric interpolation on the simplex; replacing it with a parameterized bilinear form over expert spaces might yield a smoother notion of node-to-node distance and extend the method to edge-weighted graphs.
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 / 6 minor

Summary. GraphMoRE proposes a mixture-of-experts framework for graph representation learning in which each expert is a Riemannian GNN operating in a space of a different curvature, and a gating network assigns per-node weights based on local topological encodings. The per-node weighted fusion of expert embeddings is claimed to embed the graph into a heterogeneous manifold with point-wise varying curvature. To compute distances between nodes in this mixed space, the paper introduces an alignment strategy that combines per-expert squared distances with pair-dependent Softmax weights. The model is trained by combining a downstream-task loss with an embedding-distortion loss, and is evaluated on link prediction and node classification on five real-world datasets plus synthetic graphs, reporting gains over Euclidean and Riemannian baselines, along with ablations and a distortion comparison.

Significance. If the geometric construction were rigorous, the paper would address a real limitation of single-curvature and product-manifold models: topological heterogeneity within a graph. The empirical results are consistently strong: GraphMoRE improves link prediction over all nine baselines on all five real datasets, with ten-run averages and released code, and the ablations identify each component's contribution. However, the advertised novelty—the 'heterogeneous manifold with varying curvatures at different points'—is asserted rather than mathematically established. The actual distance used in training and evaluation, Eq. (10), is a pair-dependent weighted average of squared expert distances, which is not shown to be the geodesic distance of any Riemannian manifold that the construction defines. This weakens the paper's central claim, though the empirical contribution could survive as a weighted-distance mixture model if the narrative is revised accordingly.

major comments (4)
  1. [Sec. 4.3, Eq. (9)] Equation (9) as written is ill-defined: if '·' denotes the standard inner product, then W_u · W_v is a scalar, and Softmax of a scalar is identically 1, so W_(u,v) would be a scalar, not the per-expert vector required by Eq. (10). If '·' denotes elementwise multiplication, that must be stated explicitly and the alignment rationale must be justified; currently no definition is given for how the weight vector W_(u,v) is produced. This ambiguity makes the distance in Eq. (10) ill-posed and directly affects the loss in Eq. (7), the link-prediction decoder, and Table 5.
  2. [Sec. 4.3 and Abstract] The central claim that the construction embeds the graph 'into a heterogeneous manifold with varying curvatures at different points' is not substantiated. No Riemannian metric, tangent-space identification, geodesic, or curvature tensor is defined for the space of concatenated embeddings in Eq. (8). Eq. (10) defines a pair-dependent convex combination of squared distances from different expert manifolds; for fixed weights it is a product-space metric, but the weights change with the node pair, so no single Riemannian distance function is shown to equal Eq. (10). The paper needs either to provide a concrete manifold construction (e.g., a warped product or a manifold with a position-dependent metric) and prove Eq. (10) is its distance, or to revise the abstract and Section 4.3 to describe a weighted-distance mixture model rather than a heterogeneous manifold.
  3. [Sec. 4.5 and Algorithm 1] The complexity analysis in Sec. 4.5, stated as O(|R||D| + |Ds| + K|D|), omits the pairwise loop explicitly shown in Algorithm 1 ('for each node pair (u, v) in G') and the all-pairs sum in Eq. (7). This pairwise computation is O(N^2) and dominates on large graphs, contradicting the stated linear complexity. The authors should either include the quadratic term in the analysis or describe how the distortion loss and alignment are computed (e.g., sampling pairs), since the current presentation does not support the scalability implied by the linear bound.
  4. [Sec. 5.2, Table 5 and Eq. (7)] The embedding-distortion comparison in Table 5 is partly circular: Eq. (7) defines LD, which is minimized as part of the training objective in Eq. (11), and Table 5 reports the same measure as the evaluation metric. Since the baselines (Q-GCN, MofitRGC) do not optimize LD, the fact that GraphMoRE achieves lower average distortion is expected by construction and does not independently demonstrate superior structure preservation. To support the claim, the authors should report distortion for a GraphMoRE variant trained without the distortion loss (e.g., the 'w/o distortion' ablation), or use an independent measure such as stress or a shortest-path-distance prediction task.
minor comments (6)
  1. [Sec. 2.2] There is a typo: 'etopological heterogeneity' should be 'topological heterogeneity'.
  2. [Sec. 4.5 title] The section heading 'Comlexity Analysis' is misspelled; it should be 'Complexity Analysis'.
  3. [Fig. 1 caption] The caption contains 'Rimannian', which should be 'Riemannian'.
  4. [Sec. 5.2] The text says 'Additionly' instead of 'Additionally'.
  5. [Sec. 5.1 / Appendix B] The multi-resolution sampling scale set R is never specified; the appendix mentions 'induced subgraphs of ego networks with different radius' but does not report the actual radii used. This hyperparameter should be listed for reproducibility.
  6. [Table 3] The description states GraphMoRE achieves the best results on the vast majority of datasets, but on PubMed the Q-GCN baseline has higher Weighted-F1 and Micro-F1 than all three GraphMoRE variants; the text should be more precise about this exception.

Circularity Check

1 steps flagged · score 6.0 of 10

The claimed 'lowest embedding distortion' is the training loss itself; downstream benchmark gains are independent but the headline distortion result is circular.

  1. fitted input called prediction [Sec. 4.2 Eq. (7), Sec. 4.4 Eq. (11), Sec. 5.2 'Comparison of Embedding Distortion' / Table 5]
    "LD = 1/|V|^2 Σ_{i,j∈V} |(d(i,j)/g(i,j))^2 − 1|, (7) ... The overall optimization objective consists of the downstream task objectives and minimization of embedding distortion, which can be formulated as: L = Ltask + λLD, (11) ... Obviously, GraphMoRE has the lowest average embedding distortion on all datasets."

    LD is not an external metric: it is the second term of the training objective in Eq. (11), so the gating network, expert embeddings, and alignment weights are all directly optimized to minimize this exact quantity. Table 5 then reports this same quantity as an evaluation result, comparing GraphMoRE to baselines that were not trained with LD. The 'lower embedding distortion' conclusion is therefore the optimized value of the paper's own loss, presented as evidence of superiority; it is forced by construction rather than an independent finding. The link-prediction and node-classification results remain externally grounded, so the circularity is partial.

full rationale

The downstream link-prediction and node-classification tables are genuine benchmark evaluations and are not circular. However, the paper's headline claim of 'lower distortion' (Abstract, Sec. 5.2, Table 5) reduces to the training objective: LD in Eq. (7) is minimized directly in Eq. (11), and the same quantity is then reported as the distortion comparison. Baselines are not trained with this loss, making the comparison unsurprising. This is a fitted input renamed as an evaluation outcome. The Sec. 4.3 assertion that the fused representation is 'equivalent to ... a heterogeneous manifold' is not backed by a Riemannian metric, tangent-space identification, or geodesic for the mixed space; that is an unsupported geometric claim rather than a circular derivation, so it does not add a separate circularity step. The curvature-initialization citation to Fu et al. (2021) is a design motivation with independent published content, not a load-bearing self-citation chain. Overall: one partial circularity in the distortion evaluation; central architecture and benchmark numbers are independent.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The method's central claim depends on hand-chosen curvatures, expert count, lambda, and embedding dimension, plus underspecified sampling radii. The most significant uncharged premise is that the fused embedding defines a valid manifold; the paper asserts this rather than proving it.

free parameters (5)
  • Initial curvature set C = {-3, -1, 0, 1, 3}
    Chosen by hand to give diverse experts; not learned or derived from data. It directly controls the mixture space and is dataset-independent (Appendix B.1).
  • Number of Riemannian experts K = 5
    Set to 5 for generality across datasets (Appendix B.1); the capacity of the mixture depends on this choice.
  • Loss balance coefficient lambda = Tuned per dataset from {1.0, 0.5, 0.1, 0.05, 0.01}
    Grid-searched per dataset (Appendix A.2), so it is a fitted constant that trades task loss against distortion.
  • Expert embedding dimension = 32 (Cora, Citeseer), 16 (others)
    Manually selected per dataset (Appendix B.1), affecting representational capacity.
  • Multi-resolution sampling scale set R = Not reported in manuscript
    The ego-network radii used in Eq. (4) are not listed; this is a hidden free parameter affecting local topology characterizations.
assumptions (5)
  • standard math kappa-stereographic model and its exp/log maps are valid and implemented as in Bachmann et al. (2020)
    The Riemannian expert operations in Eq. (1)-(2) rely on these definitions and their computational approximations without re-derivation.
  • domain assumption Graph sectional curvature computed from shortest-path distances (Eq. A.1) is a meaningful indicator of which constant-curvature space a node should be embedded in
    Used in Appendix A.1 to justify that datasets exhibit topological heterogeneity and that routing to different curvatures is needed.
  • domain assumption The gating network can be trained to route nodes to the 'optimal' expert spaces by minimizing embedding distortion LD
    This is the learning hypothesis behind Eq. (7); no guarantee is given that the differentiable path from LD to gating weights finds the true curvature optimum rather than a local one.
  • ad hoc to paper The weighted combination of expert embeddings and the aligned distance in Eq. (10) define a valid embedding space, i.e., the 'heterogeneous manifold' claim
    Asserted in Section 4.3 without constructing a Riemannian metric or verifying metric properties; the equation W(u,v)=Softmax(Wu.Wv) is also ambiguous.
  • domain assumption The distortion measure ((d/g)^2 - 1) averaged over all node pairs (Eq. 7) is a suitable objective for choosing embedding spaces
    Standard relative distortion is reasonable, but the paper does not justify it as the unique or best objective for routing.

how reviews work

0 comments
Cite this review

Pith. "Pith review of GraphMoRE: Mitigating Topological Heterogeneity via Mixture of Riemannian Experts." pith.science (2026). https://pith.science/paper/2F2PTLOV

@misc{pith2026241211085,
  author       = {Pith},
  title        = {Pith review of: GraphMoRE: Mitigating Topological Heterogeneity via Mixture of Riemannian Experts},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2F2PTLOV}},
  note         = {Machine review of arXiv:2412.11085}
}
read the original abstract

Real-world graphs have inherently complex and diverse topological patterns, known as topological heterogeneity. Most existing works learn graph representation in a single constant curvature space that is insufficient to match the complex geometric shapes, resulting in low-quality embeddings with high distortion. This also constitutes a critical challenge for graph foundation models, which are expected to uniformly handle a wide variety of diverse graph data. Recent studies have indicated that product manifold gains the possibility to address topological heterogeneity. However, the product manifold is still homogeneous, which is inadequate and inflexible for representing the mixed heterogeneous topology. In this paper, we propose a novel Graph Mixture of Riemannian Experts (GraphMoRE) framework to effectively tackle topological heterogeneity by personalized fine-grained topology geometry pattern preservation. Specifically, to minimize the embedding distortion, we propose a topology-aware gating mechanism to select the optimal embedding space for each node. By fusing the outputs of diverse Riemannian experts with learned gating weights, we construct personalized mixed curvature spaces for nodes, effectively embedding the graph into a heterogeneous manifold with varying curvatures at different points. Furthermore, to fairly measure pairwise distances between different embedding spaces, we present a concise and effective alignment strategy. Extensive experiments on real-world and synthetic datasets demonstrate that our method achieves superior performance with lower distortion, highlighting its potential for modeling complex graphs with topological heterogeneity, and providing a novel architectural perspective for graph foundation models.

Figures

Figures reproduced from arXiv: 2412.11085 by the authors.

Figure 1
Figure 1. The overall framework of GraphMoRE. We first design diverse Rimannian experts in terms of both the type of curva￾ture spaces and the degree of curving in space. Then we estimate the geometric properties around nodes from the perspective of local topology characterization, embed nodes into personalized mixed curvature spaces by fusing diverse Riemannian experts with the goal of minimizing embedding distortion. Finall… view at source ↗
Figure 2
Figure 2. Performance comparison on synthetic graphs with [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 38 canonical work pages

  1. [1]

    Bachmann, G.; B \'e cigneul, G.; and Ganea, O. 2020. Constant curvature graph convolutional networks. In ICML, 486--496. PMLR

    work page 2020

  2. [2]

    B \'e cigneul, G.; and Ganea, O.-E. 2019. Riemannian adaptive optimization methods. ICLR

    work page 2019

  3. [3]

    M.; Bruna, J.; LeCun, Y.; Szlam, A.; and Vandergheynst, P

    Bronstein, M. M.; Bruna, J.; LeCun, Y.; Szlam, A.; and Vandergheynst, P. 2017. Geometric deep learning: going beyond euclidean data. IEEE Signal Processing Magazine, 34(4): 18--42

    work page 2017

  4. [4]

    Chami, I.; Ying, Z.; R \'e , C.; and Leskovec, J. 2019. Hyperbolic graph convolutional neural networks. NeurIPS, 32

    work page 2019

  5. [5]

    Cho, S.; Cho, S.; Park, S.; Lee, H.; Lee, H.; and Lee, M. 2023. Curve Your Attention: Mixed-Curvature Transformers for Graph Representation Learning. arXiv preprint arXiv:2309.04082

    arXiv 2023

  6. [6]

    Di Giovanni, F.; Luise, G.; and Bronstein, M. 2022. Heterogeneous manifolds for curvature-aware graph embedding. ICLR (GTRL Workshop)

    work page 2022

  7. [7]

    Fu, X.; Gao, Y.; Wei, Y.; Sun, Q.; Peng, H.; Li, J.; and Li, X. 2024. Hyperbolic Geometric Latent Diffusion Model for Graph Generation. ICML

    work page 2024

  8. [8]

    Fu, X.; Li, J.; Wu, J.; Sun, Q.; Ji, C.; Wang, S.; Tan, J.; Peng, H.; and Philip, S. Y. 2021. ACE-HGNN: Adaptive curvature exploration hyperbolic graph neural network. In ICDM, 111--120. IEEE

    work page 2021

Show all 49 references
  1. [9]

    Ganea, O.; B \'e cigneul, G.; and Hofmann, T. 2018. Hyperbolic neural networks. NeurIPS, 31

    2018

  2. [10]

    Grover, A.; and Leskovec, J. 2016. node2vec: Scalable feature learning for networks. In SIGKDD, 855--864

    2016

  3. [11]

    Gu, A.; Sala, F.; Gunel, B.; and R \'e , C. 2019. Learning mixed-curvature representations in product spaces. In ICLR

    2019

  4. [12]

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

    2017

  5. [13]

    Han, H.; Li, J.; Huang, W.; Tang, X.; Lu, H.; Luo, C.; Liu, H.; and Tang, J. 2024. Node-wise Filtering in Graph Neural Networks: A Mixture of Experts Approach. arXiv preprint arXiv:2406.03464

    2024 arXiv

  6. [14]

    Han, Z.; Ma, Y.; Chen, P.; and Tresp, V. 2020. Dyernie: Dynamic evolution of riemannian manifold embeddings for temporal knowledge graph completion. EMNLP

    2020

  7. [15]

    Hou, Y.; Chen, X.; Zhu, H.; Liu, R.; Shi, B.; Liu, J.; Wu, J.; and Xu, K. 2024. NC2D: Novel Class Discovery for Node Classification. In CIKM, 849--859

    2024

  8. [16]

    Hu, F.; Wang, L.; Liu, Q.; Wu, S.; Wang, L.; and Tan, T. 2022. GraphDIVE: Graph Classification by Mixture of Diverse Experts. In IJCAI, 2080--2086

    2022

  9. [17]

    A.; Jordan, M

    Jacobs, R. A.; Jordan, M. I.; Nowlan, S. J.; and Hinton, G. E. 1991. Adaptive mixtures of local experts. Neural computation, 3(1): 79--87

    1991

  10. [18]

    Kim, S.; Lee, D.; Kang, S.; Lee, S.; and Yu, H. 2023. Learning topology-specific experts for molecular property prediction. In AAAI, volume 37, 8291--8299

    2023

  11. [19]

    P.; and Ba, J

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

    2014 arXiv

  12. [20]

    N.; and Welling, M

    Kipf, T. N.; and Welling, M. 2017. Semi-supervised classification with graph convolutional networks. ICLR

    2017

  13. [21]

    S.; et al

    Liu, J.; Yang, C.; Lu, Z.; Chen, J.; Li, Y.; Zhang, M.; Bai, T.; Fang, Y.; Sun, L.; Yu, P. S.; et al. 2023 a . Towards graph foundation models: A survey and beyond. arXiv preprint arXiv:2310.11829

    2023 arXiv

  14. [22]

    Liu, Q.; Nickel, M.; and Kiela, D. 2019. Hyperbolic graph neural networks. NeurIPS, 32

    2019

  15. [23]

    Liu, Z.; Yu, X.; Fang, Y.; and Zhang, X. 2023 b . Graphprompt: Unifying pre-training and downstream tasks for graph neural networks. In The web Conference, 417--428

    2023

  16. [24]

    Liu, Z.; Zhang, C.; Tian, Y.; Zhang, E.; Huang, C.; Ye, Y.; and Zhang, C. 2023 c . Fair graph representation learning via diverse mixture-of-experts. In The Web Conference, 28--38

    2023

  17. [25]

    Mao, H.; Chen, Z.; Tang, W.; Zhao, J.; Ma, Y.; Zhao, T.; Shah, N.; Galkin, M.; and Tang, J. 2024. Position: Graph Foundation Models Are Already Here. In ICML

    2024

  18. [26]

    Namata, G.; London, B.; Getoor, L.; Huang, B.; and Edu, U. 2012. Query-driven active surveying for collective classification. In 10th international workshop on mining and learning with graphs, volume 8, 1

    2012

  19. [27]

    Nickel, M.; and Kiela, D. 2017. Poincar \'e embeddings for learning hierarchical representations. NeurIPS, 30

    2017

  20. [28]

    Peng, W.; Varanka, T.; Mostafa, A.; Shi, H.; and Zhao, G. 2021. Hyperbolic deep neural networks: A survey. IEEE Transactions on pattern analysis and machine intelligence, 44(12): 10023--10044

    2021

  21. [29]

    Sen, P.; Namata, G.; Bilgic, M.; Getoor, L.; Galligher, B.; and Eliassi-Rad, T. 2008. Collective classification in network data. AI magazine, 29(3): 93--93

    2008

  22. [30]

    Shazeer, N.; Mirhoseini, A.; Maziarz, K.; Davis, A.; Le, Q.; Hinton, G.; and Dean, J. 2017. Outrageously large neural networks: The sparsely-gated mixture-of-experts layer. ICLR

    2017

  23. [31]

    Shchur, O.; Mumme, M.; Bojchevski, A.; and G \"u nnemann, S. 2018. Pitfalls of graph neural network evaluation. arXiv preprint arXiv:1811.05868

    2018 arXiv

  24. [32]

    Skopek, O.; Ganea, O.-E.; and B \'e cigneul, G. 2020. Mixed-curvature variational autoencoders. ICLR

    2020

  25. [33]

    Sun, L.; Huang, Z.; Wang, Z.; Wang, F.; Peng, H.; and Philip, S. Y. 2024. Motif-aware riemannian graph neural network with generative-contrastive learning. In AAAI, volume 38, 9044--9052

    2024

  26. [34]

    Sun, L.; Zhang, Z.; Ye, J.; Peng, H.; Zhang, J.; Su, S.; and Philip, S. Y. 2022. A self-supervised mixed-curvature graph neural network. In AAAI, volume 36, 4146--4155

    2022

  27. [35]

    Sun, X.; Cheng, H.; Li, J.; Liu, B.; and Guan, J. 2023. All in one: Multi-task prompting for graph neural networks. In SIGKDD, 2120--2131

    2023

  28. [36]

    Veli c kovi \'c , P.; Cucurull, G.; Casanova, A.; Romero, A.; Lio, P.; and Bengio, Y. 2018. Graph attention networks. ICLR

    2018

  29. [37]

    Wang, H.; Jiang, Z.; You, Y.; Han, Y.; Liu, G.; Srinivasa, J.; Kompella, R.; Wang, Z.; et al. 2023. Graph mixture of experts: Learning on large-scale graphs with explicit diversity modeling. NeurIPS, 36

    2023

  30. [38]

    N.; Wang, Z.; Nallapati, R.; Arnold, A.; Xiang, B.; Yu, P

    Wang, S.; Wei, X.; Nogueira dos Santos, C. N.; Wang, Z.; Nallapati, R.; Arnold, A.; Xiang, B.; Yu, P. S.; and Cruz, I. F. 2021. Mixed-curvature multi-relational graph neural network for knowledge graph completion. In The Web Conference, 1761--1771

    2021

  31. [39]

    Wu, S.; Cao, K.; Ribeiro, B.; Zou, J.; and Leskovec, J. 2024. GraphMETRO: Mitigating Complex Graph Distribution Shifts via Mixture of Aligned Experts. In The Thirty-eighth Annual Conference on Neural Information Processing Systems

    2024

  32. [40]

    Xiong, B.; Zhu, S.; Potyka, N.; Pan, S.; Zhou, C.; and Staab, S. 2022. Pseudo-riemannian graph convolutional networks. NeurIPS, 35: 3488--3501

    2022

  33. [41]

    Xu, Z.; Wen, S.; Wang, J.; Liu, G.; Wang, L.; Yang, Z.; Ding, L.; Zhang, Y.; Zhang, D.; Xu, J.; et al. 2022. AMCAD: adaptive mixed-curvature representation based advertisement retrieval system. In ICDE, 3439--3452. IEEE

    2022

  34. [42]

    C.; Liu, J.; King, I.; and Ying, R

    Yang, M.; Verma, H.; Zhang, D. C.; Liu, J.; King, I.; and Ying, R. 2024. Hypformer: Exploring Efficient Hyperbolic Transformer Fully in Hyperbolic Space. SIGKDD

    2024

  35. [43]

    Yang, M.; Zhou, M.; Li, Z.; Liu, J.; Pan, L.; Xiong, H.; and King, I. 2022. Hyperbolic graph neural networks: A review of methods and applications. arXiv preprint arXiv:2202.13852

    2022 arXiv

  36. [44]

    Yang, M.; Zhou, M.; Pan, L.; and King, I. 2023. hgcn: Tree-likeness modeling via continuous and discrete curvature learning. In SIGKDD, 2965--2977

    2023

  37. [45]

    Ying, C.; Cai, T.; Luo, S.; Zheng, S.; Ke, G.; He, D.; Shen, Y.; and Liu, T.-Y. 2021. Do transformers really perform badly for graph representation? NeurIPS, 34: 28877--28888

    2021

  38. [46]

    Zhang, Y.; Wang, X.; Shi, C.; Liu, N.; and Song, G. 2021. Lorentzian graph convolutional networks. In The web Conference, 1249--1261

    2021

  39. [47]

    Zhao, H.; Chen, A.; Sun, X.; Cheng, H.; and Li, J. 2024. All in one and one for all: A simple yet effective method towards cross-domain graph pretraining. In SIGKDD, 4443--4454

    2024

  40. [48]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...

  41. [49]

    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 gl...

Pith tools

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