Pith. sign in

REVIEW 3 major objections 5 minor 58 references

DG-Mamba: Robust and Efficient Dynamic Graph Structure Learning with Selective State Space Models

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

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

Pith's one-line read DG-Mamba claims to reduce dynamic graph structure learning from quadratic to linear time, O(T(|V|+|E|)), while improving robustness against adversarial attacks.

desk verdict Good empirical paper with a shaky complexity claim and two broken appendix proofs; the linear-time headline needs a real fix. read the letter →

arxiv 2412.08160 v4 pith:HIGIL26A submitted 2024-12-11 cs.LG

Haonan Yuan , Qingyun Sun , Zhaonan Wang , Xingcheng Fu , Cheng Ji , Yongjian Wang , Bo Jin , Jianxin Li This is my paper · ORCID
classification cs.LG
keywords dynamicgraphstructurelearningstatespacemodelsMambakernelizedattentionpositiverandomfeaturesadversarialrobustnesslinearcomplexityfuturelinkprediction
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

This paper claims that dynamic graph structure learning — refining the connectivity of a time series of graphs for downstream prediction — can be made both linear-time and more robust to adversarial perturbation. It proposes DG-Mamba, which replaces the quadratic pairwise attention of existing DGSL models with a kernelized message-passing operator that reuses two global sums, and it models the whole dynamic graph as a selective state-space system whose parameters are discretized by the learned cross-snapshot edges. Beyond efficiency, a self-supervised 'Principle of Relevant Information' regularizer is added to maximize the informativeness of the refined structures while minimizing redundancy, which the paper argues is what confers robustness. If the claims hold, DGSL becomes practical for long, large dynamic graphs and a stronger defensive baseline for link prediction under attack.

What carries the argument

The central object is the kernelized dynamic message-passing operator, built on Positive Random Features (PRF). The softmax attention weight between nodes u and v is replaced by an inner product of random feature maps φ(x) = (1/√m) Σ_i exp(ω_i^T x - ||x||^2/2), which by Mercer's theorem is an unbiased estimator of exp(x^T y). Because the aggregation becomes φ(Wz_u)^T Σ_v φ(Wz_v) Wz_v divided by φ(Wz_u)^T Σ_m φ(Wz_m), the two sums are computed once per node, eliminating the quadratic pairwise cost; the same kernel is queried to extract explicit intra- and inter-graph edge weights for structure refinement. The second piece is the Dynamic Graph Selective Scan, a selective state space model whose discrete parameters A = exp(ΔA), B = (ΔA)^{-1}(A-I)ΔB, and C are functions of the input and of the inter-graph adjacency, so that long-range dependencies are captured with a linear recurrent scan. The third piece is the PRI regularizer, L_PRI = H(Ĝ) + β D(Ĝ||G), decomposed into intra-graph entropy plus edge-level maximum-likelihood loss and inter-graph KL between the state-space output and the message-passing embedding.

What would settle it

Compute the actual KL divergence D(Ĝ||G) on a real dataset (e.g., COLLAB) before and after training with L_edge minimized, and check whether minimizing L_edge actually reduces the true KL divergence; if it does not, the equivalence in Proposition 2 collapses. Alternatively, measure empirical training-time scaling on a dense graph where |E| ≈ |V|^2: if time grows quadratically, the linear-complexity claim fails.

Watch

Extended reading notes

Core claim

The central discovery is that the quadratic spatio-temporal cost of dynamic graph structure learning — O(T|V|^2) from softmax edge weighting and O($T^{2}$) from step-pair attention — can be collapsed into a linear O(T(|V|+|E|)) pipeline without losing the ability to capture long-range dependencies. DG-Mamba achieves this with (1) a positive-random-feature kernel that approximates the softmax attention so that node-wise sums are computed once and reused, (2) a selective state-space scan where the transition, input, and step-size matrices are functions of the current input and the learned inter-graph adjacency, and (3) an information-theoretic regularizer that balances entropy and KL divergence to filter redundancy. The paper further demonstrates through experiments against 12 baselines that this linear model not only matches but often exceeds the accuracy of quadratic DGSL methods on future link prediction, and retains higher accuracy under both random structure/feature attacks and targeted evasion/poisoning attacks.

Load-bearing premise

The proof that the PRI regularizer equals an edge-level maximum-likelihood loss assumes that the learned and original adjacency matrices each sum to 1, but in practice they do not, so the information-theoretic grounding of the robustness mechanism is unproven.

Editorial extensions

If this is right

  • DGSL can be applied to large-scale, long-sequence dynamic graphs that previously caused out-of-memory errors in quadratic baselines; the paper shows near-linear scaling up to 8× sequence length.
  • Long-range temporal dependencies can be captured with a linear recurrent scan instead of pairwise temporal attention, so the model is not limited by the local Markov assumption.
  • The PRI regularizer provides a self-supervised way to denoise structures, reducing the average relative AUC drop under targeted attacks compared to all 12 baselines.
  • The kernelized operator makes the learned edge weights explicitly queryable, so intra- and inter-graph structures can be visualized and inspected.

Reading between the lines

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

  • The ablation shows that removing the kernelized operator slightly improves clean AUC, so there is a tunable trade-off: raising the projection dimension m of the random features approaches full attention at still-subquadratic cost.
  • If the PRI equivalence proof fails, the regularizer may still be useful as an empirical denoiser, but it should then be reinterpreted as a heuristic penalty rather than an information-theoretic bound; a properly normalized variational bound might restore the theory.
  • The same discretization trick — parameterizing a state-space model with a learned structure matrix — could transfer to other structured sequence problems, such as traffic forecasting or continuous-time graphs, where edge timings provide another natural discretization.
  • The paper reports empirical robustness but no guarantees, so a natural next test is certified robustness, e.g., randomized smoothing or a Lipschitz certificate for the learned dynamic structures.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes DG-Mamba, a dynamic graph structure learning framework that combines a kernelized message-passing operator with selective state space models and a self-supervised Principle of Relevant Information (PRI) regularizer. The headline claim is that the spatio-temporal complexity of DGSL is reduced to O(T(|V|+|E|)) (Eq. 24) while improving robustness against adversarial attacks. The evaluation covers future link prediction on COLLAB, Yelp, and ACT under non-targeted and targeted attacks, with 12 baselines, ablations, scaling experiments, and a publicly listed repository.

Significance. If the central complexity and robustness claims are correct, this is a useful contribution to dynamic graph structure learning: simultaneous linear scaling in both the spatial and temporal dimensions is stated as a first for DGSL, and the experimental suite is extensive, including clean and attacked settings, sensitivity analyses, and scaling studies. The paper also provides implementation detail and reproducibility artifacts. However, the theoretical appendix contains errors that directly affect the claimed guarantees, so the contribution is not yet established as written.

major comments (3)
  1. [§4.4, Eq. (24); Appendix B.1, “Efficient Structure Query”] The overall linear complexity claim is not supported by the described inter-graph structure query. The appendix states that cross-graph structures are initialized by computing cosine similarity between graphs with the initialized node embeddings and selecting the top-k edges where k equals |E|. Computing pairwise cosine similarity between consecutive snapshots requires Θ(T N^2 D) operations, and selecting the top-k from N^2 scores costs at least Ω(T N^2). No subquadratic construction (e.g., random features, locality-sensitive hashing, or approximate nearest neighbors) is described for this step, so the inference that this consumes O(T(|V|+|E|)) is unjustified. The authors must either supply a genuinely subquadratic algorithm for this initialization or explicitly exclude it from the complexity statement; otherwise Eq. (24) is unsupported.
  2. [Appendix B.2, Proposition 2, Eq. (B.18)] The loss-equivalence proof for the PRI regularizer is invalid. The proof assumes log Σ_{k,l} α^t_{kl} = log Σ_{k,l} α̂^t_{kl} = 0, which requires both adjacency sums to equal 1. In the manuscript, α^t is the original binary adjacency (sum = |E^t|) and α̂^t is a learned attention weight matrix whose total sum is generally not 1 either. Moreover, the conclusion that minimizing the KL divergence is equivalent to maximizing α̂^t_{uv} does not follow from Eq. (B.19) when α̂^t_{uv} ∈ [0,1] and α^t_{uv} ∈ {0,1}, because the quotient α̂^t_{uv}/α^t_{uv} equals 1 only in the boundary case. The claim that L_edge is equivalent to the divergence term is therefore unproven, and the information-theoretic interpretation of PRI should be either corrected or replaced by a direct statement that Eq. (21) is a heuristic likelihood-based regularizer.
  3. [Appendix B.1, “Empirical Analysis” after Eq. (B.11)] The empirical analysis of the approximation error bound is mathematically incorrect. The text claims that as ϵ and τ simultaneously approach 0, the fraction exp(6r/τ)/(mϵ²) approaches 0 and hence P(Δ < ϵ) approaches 1. In fact, as τ → 0 the exponential term grows without bound, and as ϵ → 0 the denominator shrinks, so the fraction diverges rather than approaching 0. The displayed inequality in Eq. (B.11) therefore provides no support for the claim that the Gumbel-Softmax kernel approximation error vanishes with high probability. This section must be corrected or removed.
minor comments (5)
  1. [§4.4 vs. Appendix A] The complexity accounting is internally inconsistent: §4.4 states the kernelized message-passing is O(T|V|) and the structure query is O(T|E|), while Appendix A lists O(2LT(|V|+|E|)D0) for the kernelized message-passing layer and O(2T(|V|+|E|)) for regularizing. These can be reconciled under O(T(|V|+|E|)) after treating dimensions and layer count as constants, but the discrepancy should be addressed so the two accounts agree.
  2. [Throughout] There are numerous typos and spelling errors, including “Lamma 1” (Appendix B.1), “Equivalance” (Appendix B.2 title), “sensitvity” (§5.6), “stuking” (§5.4), “secutive” (Appendix C.3), “gound-truth” (§4.3), and “res cross” (Figure 3 caption). These should be corrected.
  3. [Algorithm 1] The input list in Algorithm 1 names hyperparameters β1, β1, γ, λ, and µ, but the running text and experiments describe β1, β2, λ, and µ; γ is not defined or used elsewhere.
  4. [Figure 2] Figure 2 is very dense and the label “HBMSRAM” is unclear; the relationship between the hardware-aware scan and the discretized SSM parameters would be easier to follow with a cleaner diagram or a more explicit caption.
  5. [Appendix B.2, Eq. (B.15)] The equation contains stray “xxxx” padding characters that should be removed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central complexity and robustness claims are derived from independent components, and the self-citations appear only in background and baselines.

full rationale

The load-bearing derivation chain is not circular. Equation (7) rewrites the kernelized message-passing sum so that the per-node aggregation reuses shared sums, giving linear-in-|V| per-snapshot cost; this is a genuine algebraic factorization built on the external Positive Random Features result (Choromanski et al. 2020), not a restatement of the claimed complexity. The intra-graph structure query is explicitly restricted to existing edges in Appendix B.1, so its O(T(|V|+|E|)) cost follows from the query set itself. Future link prediction is scored on held-out test edges against 12 external baselines, so the empirical claims are not fitted re-statements of the training objective. The PRI objective in Definition 1 is an independently stated entropy-plus-divergence functional; its transformation to L_edge is defended by Proposition 2, whose normalization step (Eq. B.18) is questionable, but that is a proof-rigor problem rather than a circular definition, since L_edge is not identical to the KL term by construction. The paper's self-citations (e.g., DGIB, Yuan et al. 2024; Sun et al. 2022) are confined to related work and baseline comparison and do not supply the load-bearing content of the complexity or robustness claims. The clearest weakness is an omitted complexity analysis, not circularity: Appendix B.1 asserts that initializing inter-graph structures by all-pairs cosine similarity with top-k selection consumes O(T(|V|+|E|)), but selecting the top-k among N^2 pairs is quadratic without an additional subquadratic construction; this unsupported assertion affects correctness of the headline complexity but does not make any prediction equivalent to its input by construction.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The central claim rests on standard kernel approximation results, a Markovian assumption on graph dynamics, and a set of hyperparameters tuned on validation data. No new physical entities are postulated.

free parameters (6)
  • beta1 = 0.005 to 0.500 (validation grid)
    Trade-off between entropy and edge-level divergence in the intra-graph PRI loss (Eq. 20).
  • beta2 = 25 to 150 (validation grid)
    Trade-off in the inter-graph PRI loss (Eq. 22).
  • lambda = 0.010 to 0.100 (validation grid)
    Weight of the SSM sequence embedding in the combined representation (Eq. 17).
  • mu = 0.10 to 2.00 (validation grid)
    Lagrangian weight for the PRI loss in the overall objective (Eq. 23).
  • m (random feature projection dimension) = not specified in main text
    Dimension of the Positive Random Features in Eq. (8), controlling approximation quality and cost.
  • tau (Gumbel-Softmax temperature) = not specified
    Temperature in Eq. (B.2), controlling the discrete-continuous relaxation.
assumptions (4)
  • standard math Mercer's theorem and Positive Random Features approximate the softmax kernel unbiasedly (Lemma 1 from Choromanski et al., 2020)
    Used to derive Eq. (7) and the error bound in Proposition 1.
  • standard math Chebyshev's inequality
    Used in Proposition 1 to derive the approximation error probability bound.
  • domain assumption Assumption 1: dynamic graph follows a Markov chain
    Initially models local structure; later relaxed with SSM. Central to the kernelized message-passing derivation.
  • domain assumption Inter-graph structures can be initialized from cosine similarity between snapshots
    Appendix B.1: no ground-truth cross-snapshot edges exist; cosine similarity with top-k selection is used to initialize.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DG-Mamba: Robust and Efficient Dynamic Graph Structure Learning with Selective State Space Models." pith.science (2026). https://pith.science/paper/HIGIL26A

@misc{pith2026241208160,
  author       = {Pith},
  title        = {Pith review of: DG-Mamba: Robust and Efficient Dynamic Graph Structure Learning with Selective State Space Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HIGIL26A}},
  note         = {Machine review of arXiv:2412.08160}
}
read the original abstract

Dynamic graphs exhibit intertwined spatio-temporal evolutionary patterns, widely existing in the real world. Nevertheless, the structure incompleteness, noise, and redundancy result in poor robustness for Dynamic Graph Neural Networks (DGNNs). Dynamic Graph Structure Learning (DGSL) offers a promising way to optimize graph structures. However, aside from encountering unacceptable quadratic complexity, it overly relies on heuristic priors, making it hard to discover underlying predictive patterns. How to efficiently refine the dynamic structures, capture intrinsic dependencies, and learn robust representations, remains under-explored. In this work, we propose the novel DG-Mamba, a robust and efficient Dynamic Graph structure learning framework with the Selective State Space Models (Mamba). To accelerate the spatio-temporal structure learning, we propose a kernelized dynamic message-passing operator that reduces the quadratic time complexity to linear. To capture global intrinsic dynamics, we establish the dynamic graph as a self-contained system with State Space Model. By discretizing the system states with the cross-snapshot graph adjacency, we enable the long-distance dependencies capturing with the selective snapshot scan. To endow learned dynamic structures more expressive with informativeness, we propose the self-supervised Principle of Relevant Information for DGSL to regularize the most relevant yet least redundant information, enhancing global robustness. Extensive experiments demonstrate the superiority of the robustness and efficiency of our DG-Mamba compared with the state-of-the-art baselines against adversarial attacks.

Figures

Figures reproduced from arXiv: 2412.08160 by the authors.

Figure 1
Figure 1. A general paradigm of DGSL. Recently, there has been a growing research trend on en￾hancing the efficacy of DGNNs (Zhang et al. 2022; Zhu et al. 2023; Yuan et al. 2024). This includes a specific focus on im￾proving their ability to capture the intricate spatio-temporal correlations that surpass the first-order Weisfeiler-Leman (1- WL) graph isomorphism test (Xu et al. 2019b). Most of the DGNNs perform spatio-tempora… view at source ↗
Figure 2
Figure 2. The framework of DG-Mamba. (a) Kernelized message-passing mechanism learns both intra- and inter-graph weights [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Node scaling efficiency evaluation on Yelp. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (1 more)
Figure 6
Figure 6. Figure 6: Hyperparameter analysis. Before ( 30) After ( 30) Intra -Graph Structure Inter -Graph Structure 1.0 0.8 0.6 0.4 0.2 0.0 t = t = [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

58 extracted references · 33 canonical work pages

  1. [1]

    Aoki, M. 2013. State space modeling of time series. Springer Science & Business Media

    work page 2013

  2. [2]

    Behrouz, A.; and Hashemi, F. 2024. Graph mamba: Towards learning on graphs with state space models. arXiv preprint arXiv:2402.08678

    arXiv 2024

  3. [3]

    Choi, J.; Kim, H.; An, M.; and Whang, J. J. 2024. SpoT-Mamba: Learning Long-Range Dependency on Spatio-Temporal Graphs with Selective State Spaces. arXiv preprint arXiv:2406.11244

    arXiv 2024

  4. [4]

    M.; Likhosherstov, V.; Dohan, D.; Song, X.; Gane, A.; Sarlos, T.; Hawkins, P.; Davis, J

    Choromanski, K. M.; Likhosherstov, V.; Dohan, D.; Song, X.; Gane, A.; Sarlos, T.; Hawkins, P.; Davis, J. Q.; Mohiuddin, A.; Kaiser, L.; et al. 2020. Rethinking Attention with Performers. In International Conference on Learning Representations

    work page 2020

  5. [5]

    Feller, W. 1991. An introduction to probability theory and its applications, Volume 2, volume 81. John Wiley & Sons

    1991

  6. [6]

    Feng, F.; He, X.; Tang, J.; and Chua, T.-S. 2019. Graph adversarial training: Dynamically regularizing based on graph structure. IEEE Transactions on Knowledge and Data Engineering, 33(6): 2493--2504

    work page 2019

  7. [7]

    Fu, X.; Wei, Y.; Sun, Q.; Yuan, H.; Wu, J.; Peng, H.; and Li, J. 2023. Hyperbolic geometric graph representation learning for hierarchy-imbalance node classification. In Proceedings of the ACM Web Conference 2023, 460--468

    2023

  8. [8]

    Gu, A.; and Dao, T. 2023. Mamba: Linear-time sequence modeling with selective state spaces. arXiv preprint arXiv:2312.00752

    arXiv 2023

Show all 58 references
  1. [9]

    Gu, A.; Dao, T.; Ermon, S.; Rudra, A.; and R \'e , C. 2020. Hippo: Recurrent memory with optimal polynomial projections. Advances in neural information processing systems, 33: 1474--1487

    2020

  2. [10]

    Guo, S.; Lin, Y.; Wan, H.; Li, X.; and Cong, G. 2021. Learning dynamics and heterogeneity of spatial-temporal graph data for traffic forecasting. IEEE Transactions on Knowledge and Data Engineering, 34(11): 5415--5428

    2021

  3. [11]

    Han, Y.; Huang, G.; Song, S.; Yang, L.; Wang, H.; and Wang, Y. 2021. Dynamic neural networks: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(11): 7436--7456

    2021

  4. [12]

    Hochreiter, S.; and Schmidhuber, J. 1997. Long short-term memory. Neural computation, 9(8): 1735--1780

    1997

  5. [13]

    Huang, Y.; Miao, S.; and Li, P. 2024. What Can We Learn from State Space Models for Machine Learning on Graphs? arXiv preprint arXiv:2406.05815

    2024 arXiv

  6. [14]

    Jang, E.; Gu, S.; and Poole, B. 2022. Categorical Reparameterization with Gumbel-Softmax. In International Conference on Learning Representations

    2022

  7. [15]

    M.; Goel, R.; Jain, K.; Kobyzev, I.; Sethi, A.; Forsyth, P.; and Poupart, P

    Kazemi, S. M.; Goel, R.; Jain, K.; Kobyzev, I.; Sethi, A.; Forsyth, P.; and Poupart, P. 2020. Representation learning for dynamic graphs: A survey. Journal of Machine Learning Research, 21(70): 1--73

    2020

  8. [16]

    P.; and Ba, J

    Kingma, D. P.; and Ba, J. 2014. Adam: A method for stochastic optimization. In International Conference on Learning Representations

    2014

  9. [17]

    N.; and Welling, M

    Kipf, T. N.; and Welling, M. 2016. Variational graph auto-encoders. arXiv preprint arXiv:1611.07308

    2016 arXiv

  10. [18]

    N.; and Welling, M

    Kipf, T. N.; and Welling, M. 2022. Semi-Supervised Classification with Graph Convolutional Networks. In International Conference on Learning Representations

    2022

  11. [19]

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

    2019

  12. [20]

    Li, L.; Wang, H.; Zhang, W.; and Coster, A. 2024. Stg-mamba: Spatial-temporal graph learning via selective state space model. arXiv preprint arXiv:2403.12418

    2024 arXiv

  13. [21]

    L.; Zhang, G

    Li, Z. L.; Zhang, G. W.; Yu, J.; and Xu, L. Y. 2023. Dynamic graph structure learning for multivariate time series forecasting. Pattern Recognition, 138: 109423

    2023

  14. [22]

    Mercer, J. 1909. Xvi. functions of positive and negative type, and their connection the theory of integral equations. Philosophical transactions of the royal society of London. Series A, containing papers of a mathematical or physical character, 209(441-458): 415--446

    1909

  15. [23]

    Mikolov, T.; Chen, K.; Corrado, G.; and Dean, J. 2013. Efficient estimation of word representations in vector space. arXiv preprint arXiv:1301.3781

    2013 arXiv

  16. [24]

    Pareja, A.; Domeniconi, G.; Chen, J.; Ma, T.; Suzumura, T.; Kanezashi, H.; Kaler, T.; Schardl, T.; and Leiserson, C. 2020. Evolvegcn: Evolving graph convolutional networks for dynamic graphs. In Proceedings of the AAAI conference on artificial intelligence, volume 34, 5363--5370

    2020

  17. [25]

    Principe, J. C. 2010. Information theoretic learning: Renyi's entropy and kernel perspectives. Springer Science & Business Media

    2010

  18. [26]

    Rossi, E.; Chamberlain, B.; Frasca, F.; Eynard, D.; Monti, F.; and Bronstein, M. 2020. Temporal graph networks for deep learning on dynamic graphs. arXiv preprint arXiv:2006.10637

    2020 arXiv

  19. [27]

    Sankar, A.; Wu, Y.; Gou, L.; Zhang, W.; and Yang, H. 2020. Dysat: Deep neural representation learning on dynamic graphs via self-attention networks. In Proceedings of the 13th international conference on web search and data mining, 519--527

    2020

  20. [28]

    Seo, Y.; Defferrard, M.; Vandergheynst, P.; and Bresson, X. 2018. Structured sequence modeling with graph convolutional recurrent networks. In Neural Information Processing: 25th International Conference, ICONIP 2018, Siem Reap, Cambodia, December 13-16, 2018, Proceedings, Par...

    2018

  21. [29]

    Shen, Z.; Zhang, M.; Zhao, H.; Yi, S.; and Li, H. 2021. Efficient attention: Attention with linear complexities. In Proceedings of the IEEE/CVF winter conference on applications of computer vision, 3531--3539

    2021

  22. [30]

    Sun, L.; Zhang, Z.; Wang, F.; Ji, P.; Wen, J.; Su, S.; and Philip, S. Y. 2022 a . Aligning dynamic social networks: An optimization over dynamic graph autoencoder. IEEE Transactions on Knowledge and Data Engineering, 35(6): 5597--5611

    2022

  23. [31]

    Sun, Q.; Li, J.; Peng, H.; Wu, J.; Fu, X.; Ji, C.; and Philip, S. Y. 2022 b . Graph structure learning with variational information bottleneck. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 36, 4165--4174

    2022

  24. [32]

    Sun, Q.; Li, J.; Yuan, H.; Fu, X.; Peng, H.; Ji, C.; Li, Q.; and Yu, P. S. 2022 c . Position-aware structure learning for graph topology-imbalance by relieving under-reaching and over-squashing. In Proceedings of the 31st ACM International Conference on Information & Knowledge...

    2022

  25. [33]

    Tang, J.; Qian, T.; Liu, S.; Du, S.; Hu, J.; and Li, T. 2022. Spatio-temporal latent graph structure learning for traffic forecasting. In 2022 International Joint Conference on Neural Networks (IJCNN), 1--8. IEEE

    2022

  26. [34]

    Tang, J.; Wu, S.; Sun, J.; and Su, H. 2012. Cross-domain collaboration recommendation. In Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining, 1285--1293

    2012

  27. [35]

    N.; Kaiser, .; and Polosukhin, I

    Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A. N.; Kaiser, .; and Polosukhin, I. 2017. Attention is all you need. Advances in neural information processing systems, 30

    2017

  28. [36]

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

    2018

  29. [37]

    Wang, B.; Jia, J.; Cao, X.; and Gong, N. Z. 2021. Certified robustness of graph neural networks against adversarial structural perturbation. In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, 1645--1653

    2021

  30. [38]

    Wang, C.; Tsepa, O.; Ma, J.; and Wang, B. 2024. Graph-mamba: Towards long-range graph sequence modeling with selective state spaces. arXiv preprint arXiv:2402.00789

    2024 arXiv

  31. [39]

    Wei, Y.; Yuan, H.; Fu, X.; Sun, Q.; Peng, H.; Li, X.; and Hu, C. 2024. Poincar \'e Differential Privacy for Hierarchy-aware Graph Embedding. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, 9160--9168

    2024

  32. [40]

    Wu, H.; Wang, C.; Tyshetskiy, Y.; Docherty, A.; Lu, K.; and Zhu, L. 2019. Adversarial examples for graph data: deep insights into attack and defense. In Proceedings of the 28th International Joint Conference on Artificial Intelligence, 4816--4823

    2019

  33. [41]

    P.; and Yan, J

    Wu, Q.; Zhao, W.; Li, Z.; Wipf, D. P.; and Yan, J. 2022. Nodeformer: A scalable graph structure learning transformer for node classification. Advances in Neural Information Processing Systems, 35: 27387--27401

    2022

  34. [42]

    Xu, K.; Chen, H.; Liu, S.; Chen, P.-Y.; Weng, T.-W.; Hong, M.; and Lin, X. 2019 a . Topology attack and defense for graph neural networks: An optimization perspective. arXiv preprint arXiv:1906.04214

    2019 arXiv

  35. [43]

    Xu, K.; Hu, W.; Leskovec, J.; and Jegelka, S. 2019 b . How Powerful are Graph Neural Networks? In International Conference on Learning Representations

    2019

  36. [44]

    Yuan, H.; Sun, Q.; Fu, X.; Ji, C.; and Li, J. 2024. Dynamic Graph Information Bottleneck. In Proceedings of the ACM on Web Conference 2024, 469--480

    2024

  37. [45]

    Zhang, H.; Han, X.; Xiao, X.; and Bai, J. 2023 a . Time-aware Graph Structure Learning via Sequence Prediction on Temporal Graphs. In Proceedings of the 32nd ACM International Conference on Information and Knowledge Management, 3288--3297

    2023

  38. [46]

    Zhang, M.; Wu, S.; Yu, X.; Liu, Q.; and Wang, L. 2022. Dynamic graph neural networks for sequential recommendation. IEEE Transactions on Knowledge and Data Engineering, 35(5): 4741--4753

    2022

  39. [47]

    Zhang, Q.; Chang, J.; Meng, G.; Xiang, S.; and Pan, C. 2020. Spatio-temporal graph structure learning for traffic forecasting. In Proceedings of the AAAI conference on artificial intelligence, volume 34, 1177--1185

    2020

  40. [48]

    Zhang, S.; Xiong, Y.; Zhang, Y.; Sun, Y.; Chen, X.; Jiao, Y.; and Zhu, Y. 2023 b . RDGSL: Dynamic Graph Representation Learning with Structure Learning. In Proceedings of the 32nd ACM International Conference on Information and Knowledge Management, 3174--3183

    2023

  41. [49]

    Zhang, X.; and Zitnik, M. 2020. Gnnguard: Defending graph neural networks against adversarial attacks. Advances in neural information processing systems, 33: 9263--9275

    2020

  42. [50]

    Zhang, Z.; Lin, H.; and Gao, Y. 2018. Dynamic hypergraph structure learning. In Proceedings of the 27th International Joint Conference on Artificial Intelligence, 3162--3169

    2018

  43. [51]

    Zhou, Z.; Yang, K.; Liang, Y.; Wang, B.; Chen, H.; and Wang, Y. 2023. Predicting collective human mobility via countering spatiotemporal heterogeneity. IEEE Transactions on Mobile Computing

    2023

  44. [52]

    Zhu, D.; Zhang, Z.; Cui, P.; and Zhu, W. 2019. Robust graph convolutional networks against adversarial attacks. In Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining, 1399--1407

    2019

  45. [53]

    Zhu, Y.; Cong, F.; Zhang, D.; Gong, W.; Lin, Q.; Feng, W.; Dong, Y.; and Tang, J. 2023. Wingnn: Dynamic graph neural networks with random gradient aggregation window. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 3650--3662

    2023

  46. [54]

    Zhu, Y.; Xu, W.; Zhang, J.; Liu, Q.; Wu, S.; and Wang, L. 2021. Deep graph structure learning for robust representations: A survey. arXiv preprint arXiv:2103.03036, 14: 1--1

    2021 arXiv

  47. [55]

    u gner, D.; Akbarnejad, A.; and G \

    Z \"u gner, D.; Akbarnejad, A.; and G \"u nnemann, S. 2018. Adversarial attacks on neural networks for graph data. In Proceedings of the 24th ACM SIGKDD international conference on knowledge discovery & data mining, 2847--2856

    2018

  48. [56]

    u gner, D.; and G \

    Z \"u gner, D.; and G \"u nnemann, S. 2019. Certifiable robustness and robust training for graph convolutional networks. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 246--256

    2019

  49. [57]

    , " * 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...

  50. [58]

    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.