Pith. sign in

REVIEW 3 major objections 6 minor 53 references

Stability of Flow Models for Graph Signals

T0 review · 3 major / 6 minor · reviewed 2026-07-09 · glm-5.2

Pith's one-line read Graph signal generators stay stable under topology noise

desk verdict Wasserstein stability bounds for GNN-parametrized continuous normalizing flows on graphs — sound theory with a fixable proof gap and a training-implementation mismatch worth flagging. read the letter →

arxiv 2607.07510 v1 pith:NJVIPWND submitted 2026-07-08 eess.SP cs.AIcs.LG

classification eess.SPcs.AIcs.LG
keywords graphneuralnetworkscontinuousnormalizingflowsWassersteinstabilitypermutationequivarianceflowmatchingsignalprocessingrelativeperturbationsLipschitzregularization
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

When a generative flow model uses a graph neural network as its vector field, the resulting probability distribution over graph signals degrades gracefully under structural perturbations of the graph. Specifically, for a relative perturbation of magnitude ε, the 2-Wasserstein distance between generated distributions on the nominal and perturbed graphs is bounded by a product of three factors: Γ (inherited GNN stability), Ω_t (exponential error growth governed by the one-sided Lipschitz constant of the flow dynamics), and C'_g (a trajectory-dependent supremum). This bound holds for both the continuous-time ODE and its discrete Euler or Heun approximations. The key structural insight is that the spatial Lipschitz constant M of the vector field drives the exponential growth factor Ω_t, making it the dominant controllable source of instability. Penalizing M during training—via a spectral bound on the GNN filter responses—yields models that are more robust to graph estimation errors without degrading generation quality.

What carries the argument

The Wasserstein stability bound (Theorem 1, Corollary 2) for the continuous flow, its discrete counterparts (Theorem 2, Corollaries 3–4) for Euler and Heun samplers, and the spectral Lipschitz bound (Proposition 3) that enables graph-aware regularization of the vector field during flow matching training.

What would settle it

Train a GNN-parametrized flow where the learned filter coefficients produce a frequency response h(λ) that is not integral Lipschitz, then measure the empirical Wasserstein distance between generated distributions on nominal and perturbed graphs; if the distance grows superlinearly in ε or violates the predicted Ω_t scaling, the bound's dependence on the inherited GNN stability assumption is exposed.

Watch

Extended reading notes

Core claim

The paper proves that a continuous normalized flow parameterized by a GNN inherits permutation equivariance for both the continuous ODE and discrete samplers (Euler, Heun), and derives explicit Wasserstein stability bounds showing that distributional drift under relative graph perturbations factors into (i) base GNN stability Γ, (ii) an exponential growth term Ω_t governed by the one-sided Lipschitz constant m_t of the vector field, and (iii) a trajectory-dependent constant C'_g. The exponential growth factor is the mechanism through which small per-step vector field errors accumulate over the integration horizon, and because Ω_t is monotone in the Lipschitz constant M, regularizing M via a谱

Load-bearing premise

The entire stability framework inherits a bound from prior work requiring that the GNN's graph filter frequency response be integral Lipschitz (constant C) and uniformly bounded (constant B), but the training procedure only regularizes the spatial Lipschitz constant M, leaving C and B uncontrolled—if the learned filters violate integral Lipschitzness, the base stability bound does not hold.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. The paper studies permutation equivariance and stability of continuous normalized flow (CNF) models parameterized by graph neural networks (GNNs) for graph signal generation. The central contribution is a Wasserstein stability bound (Theorem 1, Corollary 2) showing that under relative graph perturbations of magnitude ε, the 2-Wasserstein distance between generated distributions on nominal and perturbed graphs is bounded by Ω_t · C'_g · (Γε + O(ε²)), where Γ captures base GNN stability, Ω_t is an exponential growth factor from the one-sided Lipschitz constant, and C'_g is a trajectory-dependent supremum. The framework extends to discrete Euler and Heun samplers (Theorem 2, Corollaries 3–4). Motivated by the bounds, the authors propose a regularized flow matching (RFM) objective penalizing the spatial Lipschitz constant M, and validate on synthetic SBM graphs and fMRI connectome data.

Significance. The paper provides a clean, self-contained theoretical treatment of a timely problem: how structural perturbations propagate through generative flow dynamics on graphs. The proofs use standard tools appropriately (Grönwall's inequality, triangle inequality, induction). The derivation is from first principles, building on the external GNN stability result of Gama et al. [30] without circularity. The practical regularization strategy is directly motivated by the theory, and the experiments honestly report that the bounds are loose but that bound-informed regularization still improves robustness. The extension to discrete samplers with explicit Euler and Heun bounds, including the convergence to the continuous bound as h→0, is a solid contribution.

major comments (3)
  1. Appendix B, proof of Theorem 1: The one-sided Lipschitz condition (13) is stated for the nominal graph S, but in the proof it is applied to u_t^θ(·; S̃) on the perturbed graph S̃. Specifically, the bound ⟨x̃_t - P₀x_t, u_t^θ(x̃_t; S̃) - u_t^θ(P₀x_t; S̃)⟩ ≤ m_t‖x̃_t - P₀x_t‖² invokes (13) with the perturbed GSO S̃, which is not covered by the assumption as written. The same pattern recurs in Appendix G (proof of Theorem 2), where the state stability condition (Definition 4) is defined for S but applied to S̃ in the recurrence ‖T_k^θ(x̃_k; S̃) - T_k^θ(P₀x_k; S̃)‖ ≤ α‖x̃_k - P₀x_k‖. This is a load-bearing gap: the proof structure can be repaired (e.g., by decomposing u_t^θ(x; S̃) - u_t^θ(y; S̃) into nominal-graph and cross-graph terms, applying (13) to the former and (12) to the latter, introducing additional O(Γε) contributions that change the constant prefactor but preserve the bound's O(
  2. §II-B and §V: The base GNN stability bound (4) from [30, Thm. 4] requires the filter frequency response h(λ) to be integral Lipschitz with constant C and uniformly bounded by B. These conditions constrain the learned filter coefficients but are not enforced during training—the paper only regularizes the spatial Lipschitz constant M (Proposition 3, Eq. 25), leaving C and B uncontrolled. The paper acknowledges this in §VII ('exploring...jointly regularizing the spatial and frequency-response Lipschitz constants') but does not state as a formal assumption that the trained filters satisfy integral Lipschitzness, nor does it verify this empirically. If the learned filters violate these conditions, the base bound (4) does not hold and the entire stability edifice (Theorems 1–2, Corollaries 1–4) is unsupported. The authors should either (a) add an explicit assumption that trained filters are验证d
  3. §V, Eqs. (22)–(24) and Proposition 3: The bound on M in (24) is stated as M ≤ ∏_ℓ max_i ‖∑_p Θ_ℓp λ_i^p‖_2, but the proof in Appendix J derives M ≤ ∏_ℓ M_ℓ where M_ℓ := max_i ‖∑_p Θ_ℓp λ_i^p‖_2. The product of maxima is not the same as the maximum of products. Please clarify whether (24) is the tight bound from the proof or a further relaxation, and if the latter, note this explicitly.
minor comments (6)
  1. §III, Eq. (12): The bound ‖g(x,t)‖ appears on the right-hand side, but the subsequent definition of C_g(x₀) in Theorem 1 evaluates ‖g(Φ_τ^θ(P₀^⊤x₀; S), τ)‖ along the trajectory. Please clarify that (12) is being applied with x = Φ_τ^θ(x₀; S̃) (the perturbed trajectory) while C_g is defined along the unperturbed trajectory, and justify that the latter bounds the former.
  2. §VI-A: The architecture description states F₀ = 65 (input feature dimension from 1 node feature + 64 time embedding dimensions), but earlier in §III it is assumed F₀ = F_L = 1. Please reconcile this discrepancy or note that the theoretical analysis assumes d=1 for simplicity while experiments use d=65.
  3. Figures 1–2: The y-axis labels on the center panels ('K(x₀; S)') are not defined in the text. Presumably this refers to the empirical stability metric ‖Ψ_k^θ(x₀; S̃) - Ψ_k^θ(x₀; S)‖, but this should be stated explicitly.
  4. §VI-A, Synthetic perturbation: The perturbation model generates a random diagonal matrix E with entries drawn from [(1-ε)ε, ε]. This differs from the relative perturbation model in (3) where E is a general symmetric matrix. Please clarify the relationship or justify the restriction to diagonal E.
  5. §II-B: The eigenvector misalignment parameter δ is stated as δ ≤ 8 but the source [30, Thm. 1] should be checked for the exact conditions under which this bound holds.
  6. Typos: 'errrors' (§II-D), 'F(6)' should likely be 'g (6)' in Appendix H (Euler sampler graph stability verification).

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for a careful and constructive reading of our manuscript. The report identifies three substantive issues, all of which are valid and warrant revision. We address each below.

read point-by-point responses
  1. Referee: Appendix B, proof of Theorem 1: The one-sided Lipschitz condition (13) is stated for the nominal graph S, but in the proof it is applied to u_t^θ(·; S̃) on the perturbed graph S̃. The same pattern recurs in Appendix G (proof of Theorem 2), where Definition 4 is defined for S but applied to S̃. This is a load-bearing gap.

    Authors: The referee is correct. Condition (13) is stated for the nominal GSO S, but in the proof of Theorem 1 (Appendix B) it is applied to u_t^θ(·; S̃), i.e., the vector field evaluated on the perturbed graph. The same issue arises in Appendix G for Definition 4. As the referee notes, the proof structure can be repaired by decomposing u_t^θ(x̃_t; S̃) - u_t^θ(P₀x_t; S̃) into a nominal-graph term (to which (13) applies directly) and a cross-graph perturbation term (bounded via (12)). This introduces an additional O(Γε) contribution that changes the constant prefactor but preserves the O(Γε + O(ε²)) structure of the bound. We will revise the assumption statements in (13) and Definition 4 to explicitly require that the one-sided Lipschitz and state stability conditions hold for all GSOs in the relative perturbation neighborhood of S (i.e., for any S̃ satisfying (3) with ||E||₂ ≤ ε). This is a natural and mild strengthening: the conditions depend on the filter coefficients Θ_{ℓp}, which are shared across S and S̃, and on the spectrum of the GSO, which changes continuously under relative perturbations. We will also add a remark explaining why this strengthened assumption is satisfied under the same integral Lipschitz and boundedness conditions on the filter frequency response that underpin the base GNN stability bound (4). The revised proofs in Appendices B and G will explicitly show the decomposition the referee suggests. revision: yes

  2. Referee: §II-B and §V: The base GNN stability bound (4) from [30, Thm. 4] requires the filter frequency response h(λ) to be integral Lipschitz with constant C and uniformly bounded by B. These conditions constrain the learned filter coefficients but are not enforced during training—the paper only regularizes the spatial Lipschitz constant M, leaving C and B uncontrolled. If the learned filters violate these conditions, the base bound (4) does not hold and the entire stability edifice is unsupported.

    Authors: The referee is correct that the integral Lipschitz condition (constant C) and the uniform boundedness condition (constant B) on the filter frequency response are necessary assumptions for the base GNN stability bound (4) and hence for our Theorems 1–2 and Corollaries 1–4. In the current manuscript, these are invoked as assumptions inherited from [30] but are not stated as explicit assumptions on the trained filters, nor are they verified empirically. This is a gap. We will revise the manuscript as follows. First, we will add an explicit assumption (in §II-B or §III-B) stating that the trained filter coefficients satisfy the integral Lipschitz and boundedness conditions, making clear that our stability guarantees are conditional on this assumption. Second, we will add empirical verification in §VI: after training, we will compute the bounds on C and B from (22)–(23) for the learned filters and report whether they are finite and well-behaved. We expect this to be the case given the shallow architecture (L=2, P=4) and the regularized training, but the referee is right that this must be verified rather than assumed. Third, we will strengthen the discussion in §V and §VII to acknowledge that jointly regularizing M, C, and B is the principled approach, and that our current focus on M alone is a practical simplification whose validity depends on the trained filters happening to satisfy the C and B conditions. We note that our framework is general (as stated after (4): 'our analytical framework is general and naturally extends to any base GNN satisfying a stability condition like (4)'), so the theoretical results hold for any GNN whose filters satisfy these conditions—whether by architecture design, by regularization, or by verification. revision: yes

  3. Referee: §V, Eqs. (22)–(24) and Proposition 3: The bound on M in (24) is stated as M ≤ ∏_ℓ max_i ‖∑_p Θ_ℓp λ_i^p‖_2, but the proof in Appendix J derives M ≤ ∏_ℓ M_ℓ where M_ℓ := max_i ‖∑_p Θ_ℓp λ_i^p‖_2. The product of maxima is not the same as the maximum of products. Please clarify whether (24) is the tight bound from the proof or a further relaxation.

    Authors: The referee is correct that there is a notational discrepancy between (24) and the proof in Appendix J. The proof derives M ≤ ∏_{ℓ=1}^{L} M_ℓ, where M_ℓ := max_i ‖∑_p Θ_{ℓp} λ_i^p‖_2. This is indeed a product of per-layer maxima, not a maximum of products. Equation (24) as written, M ≤ ∏_ℓ max_i ‖∑_p Θ_{ℓp} λ_i^p‖_2, is in fact the same expression—the product symbol ∏_ℓ ranges over layers and the max_i is taken within each layer. So (24) is the tight bound from the proof, not a further relaxation. However, we agree the notation is ambiguous as written because the scope of the product and maximum operators is not clearly delineated. We will revise (24) to make the notation unambiguous, writing M ≤ ∏_{ℓ=1}^{L} M_ℓ with M_ℓ := max_{i∈{1,...,N}} ‖∑_{p=0}^{P-1} Θ_{ℓp} λ_i^p‖_2, and we will add a sentence clarifying that this is a product of per-layer maxima (each maximum taken over graph frequencies), which is the direct result of the recursive application of the layer-wise Lipschitz bound in Appendix J. revision: partial

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: stability bounds derived from first principles via Grönwall's inequality, using an external GNN stability result from non-overlapping authors as input.

full rationale

The paper's central stability bounds (Theorems 1–2, Corollaries 1–4) are derived from first principles by applying Grönwall's inequality to the ODE error dynamics. The key input is the GNN stability bound (4) from [30] (Gama, Bruna, Ribeiro—none of whom are authors of this paper), which is an externally established result. The regularization term (Proposition 3) is derived within the paper from the GNN architecture's spectral properties, not fitted to the target outcome. The 'prediction' that regularizing the spatial Lipschitz constant M improves robustness is tested empirically against external benchmarks (SBM and fMRI data), not forced by construction. The proof gap flagged by the skeptic (applying the one-sided Lipschitz condition (13), stated for the nominal graph S, to the perturbed graph S̃ in Appendix B) is a correctness concern, not a circularity issue—it does not make the result equivalent to its inputs by definition. No self-definitional, fitted-input, or self-citation circularity is present in the derivation chain.

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

No new entities, particles, forces, or dimensions are introduced. The paper works entirely within the existing frameworks of GNNs, continuous normalizing flows, and graph signal processing.

free parameters (5)
  • μ = 0.01
    Regularization weight selected via grid search (§VI-A). Controls the trade-off between flow matching loss and Lipschitz penalty.
  • K = 100
    Number of Euler discretization steps. Chosen, not derived.
  • L = 2
    GNN depth. Architectural choice.
  • F = 4
    GNN hidden width. Architectural choice.
  • P = 4
    Number of filter taps. Architectural choice.
assumptions (5)
  • domain assumption Base GNN satisfies stability bound (4) from [30, Thm. 4], requiring integral Lipschitz frequency response h(λ) with constant C and uniform bound ‖H(Λ)‖₂ ≤ B
    Invoked in §II-B and used in §III-B via equation (12) to bound vector field perturbation error. This is the foundational input to all stability results.
  • domain assumption Vector field is Lipschitz continuous in state variable with constant M (equation 9)
    Stated in §III as assumption for Picard-Lindelöf existence/uniqueness and used throughout the stability analysis.
  • domain assumption Relative perturbation model (3): perturbed GSO satisfies S̃ = S + ½(ES + SE) modulo permutation
    Defined in §II-B following [30]. All stability bounds are derived under this specific perturbation model.
  • domain assumption One-sided Lipschitz condition (13): ⟨u(x;S) - u(y;S), x-y⟩ ≤ m_t‖x-y‖²
    Invoked in Theorem 1 to obtain the integral form of the stability bound. Standard for ODE stability analysis.
  • standard math Grönwall's inequality
    Used in Appendix B to derive the exponential growth factor Ω_t from the differential inequality.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Stability of Flow Models for Graph Signals." pith.science (2026). https://pith.science/paper/NJVIPWND

@misc{pith2026260707510,
  author       = {Pith},
  title        = {Pith review of: Stability of Flow Models for Graph Signals},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NJVIPWND}},
  note         = {Machine review of arXiv:2607.07510}
}
read the original abstract

Generating signals on graphs requires permutation-equivariant models that exhibit stability with respect to relative structural perturbations. While favorable stability properties of Graph Neural Networks (GNNs) have been well documented, it is unclear how structural errors propagate through the dynamics of continuous generative flow models that are gaining traction for graph signal generation. In this paper, we analyze continuous normalized flow models parameterized by GNNs and show that permutation equivariance is preserved for both the resulting continuous-time ordinary differential equations and their discrete numerical approximations used as graph signal samplers. Our primary contribution is to derive explicit stability bounds on the generated probability distributions, which quantify how relative graph perturbations affect the final sampled signals. Motivated by these theoretical bounds, we introduce a stability-promoting regularized flow matching strategy that actively penalizes the spatial Lipschitz constant of the vector field during model training. Experiments using synthetic smooth signals on stochastic block model graphs and real-world fMRI signals on brain connectomes demonstrate that this bound-oriented approach yields generative models that are more robust to structural noise, without sacrificing output quality.

Figures

Figures reproduced from arXiv: 2607.07510 by the authors.

Figure 1
Figure 1. Performance and stability on a SBM graph with smooth signals under a controlled synthetic perturbation of magnitude [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. Performance and stability on fMRI signals over a functional brain connectome under data-driven graph perturbations. The x-axis represents the ratio [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 53 canonical work pages

  1. [30]

    Stability properties of graph neural networks,

    F. Gama, J. Bruna, and A. Ribeiro, “Stability properties of graph neural networks,”IEEE Trans. Signal Process., vol. 68, p. 5680–5695, 2020

  2. [1]

    Flow matching for generative modeling,

    Y . Lipman, R. T. Chen, H. Ben-Hamu, M. Nickel, and M. Le, “Flow matching for generative modeling,” inProc. Int. Conf. Learn. Represen- tations, 2023, pp. 1–28

  3. [2]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” inProc. Adv. Neural. Inf. Process. Syst., 2020, pp. 1–12

  4. [3]

    Neural ordinary differential equations,

    R. T. Q. Chen, Y . Rubanova, J. Bettencourt, and D. Duvenaud, “Neural ordinary differential equations,” inProc. Adv. Neural. Inf. Process. Syst., 2018, pp. 1–13

  5. [4]

    Graph normalizing flows,

    J. Liu, A. Kumar, J. Ba, J. Kiros, and K. Swersky, “Graph normalizing flows,” inProc. Adv. Neural. Inf. Process. Syst., 2019, pp. 1–11

  6. [5]

    Permu- tation invariant graph generation via score-based generative modeling,

    C. Niu, Y . Song, J. Song, S. Zhao, A. Grover, and S. Ermon, “Permu- tation invariant graph generation via score-based generative modeling,” inProc. Int. Conf. Artif. Intell. Statist., 2020, pp. 1–10

  7. [6]

    Score-based generative modeling of graphs via the system of stochastic differential equations,

    J. Jo, S. Lee, and S. J. Hwang, “Score-based generative modeling of graphs via the system of stochastic differential equations,” inProc. Int. Conf. Mach. Learn., 2022, pp. 10 362–10 383

  8. [7]

    Generative diffusion models on graphs: Methods and applications,

    C. Liu, W. Fan, Y . Liu, J. Li, H. Li, H. Liu, J. Tang, and Q. Li, “Generative diffusion models on graphs: Methods and applications,” in Proc. Int. Joint Cont. Artificial Intell., 2023, pp. 6702–6711

Show all 53 references
  1. [8]

    Digress: Discrete denoising diffusion for graph generation,

    C. Vignac, I. Krawczuk, A. Siraudin, B. Wang, V . Cevher, and P. Frossard, “Digress: Discrete denoising diffusion for graph generation,” inProc. Int. Conf. Learn. Representations, 2023, pp. 1–22

  2. [9]

    Prior-informed flow matching for graph reconstruction,

    H. Chen, N. Zilberstein, and S. Segarra, “Prior-informed flow matching for graph reconstruction,” inICML Wrksp. Struct. Prob. Inference and Gen. Modeling, 2026, pp. 1–33

  3. [10]

    Graph-aware diffusion for signal generation,

    S. Rozada, Vimal K B, A. Cavallo, A. G. Marques, H. Jamali-Rad, and E. Isufi, “Graph-aware diffusion for signal generation,” inProc. IEEE Intl. Conf. Acoustics, Speech and Signal Process. (ICASSP), 2026, pp. 461–465

  4. [11]

    Graph signal generative diffusion models,

    Y . B. Uslu, S. Hadou, S. Rozada, S. S. Bidokhti, and A. Ribeiro, “Graph signal generative diffusion models,” inProc. IEEE Intl. Conf. Acoustics, Speech and Signal Process. (ICASSP), 2026, pp. 626–630

  5. [12]

    Graph signal diffusion model for collaborative filtering,

    Y . Zhu, C. Wang, Q. Zhang, and H. Xiong, “Graph signal diffusion model for collaborative filtering,” inProc. ACM SIGIR Conf. Res. Develop. Inf. Retrieval, 2024, pp. 1380–1390

  6. [13]

    Topological schr ¨odinger bridge matching,

    M. Yang, “Topological schr ¨odinger bridge matching,” inProc. Int. Conf. Learn. Representations, 2025, pp. 1–42

  7. [14]

    Graph signal processing: Overview, challenges, and applications,

    A. Ortega, P. Frossard, J. Kova ˇcevi´c, J. M. Moura, and P. Vandergheynst, “Graph signal processing: Overview, challenges, and applications,”Proc. IEEE, vol. 106, no. 5, pp. 808–828, 2018

  8. [15]

    A graph signal processing perspective on functional brain imaging,

    W. Huang, T. A. Bolton, J. D. Medaglia, D. S. Bassett, A. Ribeiro, and D. Van De Ville, “A graph signal processing perspective on functional brain imaging,”Proc. IEEE, vol. 106, no. 5, pp. 868–885, 2018

  9. [16]

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

    B. Yu, H. Yin, and Z. Zhu, “Spatio-temporal graph convolutional networks: A deep learning framework for traffic forecasting,” inProc. Int. Joint Cont. Artificial Intell., 2018, pp. 3634–3640

  10. [17]

    Graph signal processing in applications to sensor networks, smart grids, and smart cities,

    I. Jabło ´nski, “Graph signal processing in applications to sensor networks, smart grids, and smart cities,”IEEE Sensors J., vol. 17, no. 23, pp. 7659–7666, 2017

  11. [18]

    Grid-graph signal processing (Grid- GSP): A graph signal processing framework for the power grid,

    R. Ramakrishna and A. Scaglione, “Grid-graph signal processing (Grid- GSP): A graph signal processing framework for the power grid,”IEEE Trans. Signal Process., vol. 69, pp. 2725–2739, 2021

  12. [19]

    Graph sig- nal diffusion models for wireless resource allocation,

    Y . B. Uslu, S. Hadou, S. S. Bidokhti, and A. Ribeiro, “Graph sig- nal diffusion models for wireless resource allocation,” 2026; see also arXiv:2604.05175 [eess.SP]

  13. [20]

    Data augmentation in classification and segmentation: A survey and new strategies,

    K. Alomar, H. I. Aysel, and X. Cai, “Data augmentation in classification and segmentation: A survey and new strategies,”J. Imaging, vol. 9, no. 2, pp. 1–26, 2023

  14. [21]

    Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations,

    M. Raissi, P. Perdikaris, and G. E. Karniadakis, “Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations,”J. Comput. Phys., vol. 378, pp. 686–707, 2019

  15. [22]

    Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges,

    M. M. Bronstein, J. Bruna, T. Cohen, and P. Veli ˇckovi´c, “Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges,”arXiv preprint arXiv:2104.13478, 2021

  16. [23]

    Intriguing properties of neural networks,

    C. Szegedy, W. Zaremba, I. Sutskever, J. Bruna, D. Erhan, I. Goodfellow, and R. Fergus, “Intriguing properties of neural networks,” inProc. Int. Conf. Learn. Representations, 2014, pp. 1–10

  17. [24]

    Stability and generalization,

    O. Bousquet and A. Elisseeff, “Stability and generalization,”J. Mach. Learn. Res., vol. 2, pp. 499–526, 2002

  18. [25]

    Adversarial attacks on neural networks for graph data,

    D. Z ¨ugner, A. Akbarnejad, and S. G ¨unnemann, “Adversarial attacks on neural networks for graph data,”Intl. Conf. Knowledge Discovery Data Mining, pp. 2847–2856, 2018

  19. [26]

    E. D. Kolaczyk,Statistical Analysis of Network Data: Methods and Models. New York, NY: Springer, 2009

  20. [27]

    On the use of correlation as a measure of network connectivity,

    A. Zalesky, A. Fornito, and E. Bullmore, “On the use of correlation as a measure of network connectivity,”Neuroimage, vol. 60, no. 4, pp. 2096–2106, 2012

  21. [28]

    Connecting the dots: Identifying network structure via graph signal processing,

    G. Mateos, S. Segarra, A. G. Marques, and A. Ribeiro, “Connecting the dots: Identifying network structure via graph signal processing,”IEEE Signal Process. Mag., vol. 36, no. 3, pp. 16–43, May 2019

  22. [29]

    Learning graphs from data: A signal representation perspective,

    X. Dong, D. Thanou, M. Rabbat, and P. Frossard, “Learning graphs from data: A signal representation perspective,”IEEE Signal Process. Mag., vol. 36, no. 3, pp. 44–63, 2019

  23. [31]

    Transferability properties of graph neural networks,

    L. Ruiz, L. F. O. Chamon, and A. Ribeiro, “Transferability properties of graph neural networks,”IEEE Trans. Signal Process., vol. 71, pp. 3474 – 3489, 2023

  24. [32]

    Interpretable stability bounds for spectral graph filters,

    H. Kenlay, D. Thanou, and X. Dong, “Interpretable stability bounds for spectral graph filters,” inProc. Int. Conf. Mach. Learn., 2021, pp. 5388–5397

  25. [33]

    Trans- ferability of spectral graph convolutional neural networks,

    R. Levie, W. Huang, L. Bucci, M. Bronstein, and G. Kutyniok, “Trans- ferability of spectral graph convolutional neural networks,”J. Mach. Learn. Res., vol. 22, pp. 1–59, 2021

  26. [34]

    Equivariant flows: Exact likelihood generative learning for symmetric densities,

    J. K ¨ohler, L. Klein, and F. No ´e, “Equivariant flows: Exact likelihood generative learning for symmetric densities,” inProc. Int. Conf. Mach. Learn., 2020, pp. 5361–5370

  27. [35]

    Graph filters for signal processing and machine learning on graphs,

    E. Isufi, F. Gama, D. I. Shuman, and S. Segarra, “Graph filters for signal processing and machine learning on graphs,”IEEE Trans. Signal Process., vol. 72, pp. 4745–4781, 2024

  28. [36]

    Convolutional neural network architectures for signals supported on graphs,

    F. Gama, A. G. Marques, G. Leus, and A. Ribeiro, “Convolutional neural network architectures for signals supported on graphs,”IEEE Trans. Signal Process., vol. 67, no. 4, p. 1034–1049, 2019

  29. [37]

    Semi-supervised classification with graph convolutional networks,

    T. N. Kipf and M. Welling, “Semi-supervised classification with graph convolutional networks,” inProc. Int. Conf. Learn. Representations, 2017, pp. 1–14

  30. [38]

    Convolutional neural networks on graphs with fast localized spectral filtering,

    M. Defferrard, X. Bresson, and P. Vandergheynst, “Convolutional neural networks on graphs with fast localized spectral filtering,” inProc. Adv. Neural. Inf. Process. Syst., 2016, pp. 1–9

  31. [39]

    Graph neural networks: Architec- tures, stability and transferability,

    L. Ruiz, F. Gama, and A. Ribeiro, “Graph neural networks: Architec- tures, stability and transferability,”Proc. IEEE, vol. 109, no. 5, pp. 660– 682, 2021

  32. [40]

    Score-based generative modeling through stochastic differ- ential equations,

    Y . Song, J. Sohl-Dickstein, D. P. Kingma, A. Kumar, S. Ermon, and B. Poole, “Score-based generative modeling through stochastic differ- ential equations,” inProc. Int. Conf. Learn. Representations, 2021, pp. 1–36

  33. [41]

    Improved denoising diffusion probabilis- tic models,

    A. Q. Nichol and P. Dhariwal, “Improved denoising diffusion probabilis- tic models,” inProc. Int. Conf. Mach. Learn., 2021, pp. 8162–8171

  34. [42]

    Villaniet al.,Optimal transport: old and new

    C. Villaniet al.,Optimal transport: old and new. Springer, 2009, vol. 338

  35. [43]

    J. C. Butcher,Numerical Methods for Ordinary Differential Equations. John Wiley & Sons, 2016

  36. [44]

    Robust graph neural networks via probabilistic lipschitz constraints,

    R. Arghal, E. Lei, and S. S. Bidokhti, “Robust graph neural networks via probabilistic lipschitz constraints,”Proc. Mach. Learn. Res., vol. 168, pp. 1–13, 2022

  37. [45]

    Training stable graph neural networks through constrained learning,

    J. Cervi ˜no, L. Ruiz, and A. Ribeiro, “Training stable graph neural networks through constrained learning,” inProc. IEEE Intl. Conf. Acoustics, Speech and Signal Process. (ICASSP), 2022, pp. 4223–4227

  38. [46]

    Human Connectome Project, https://www.humanconnectome.org/

  39. [47]

    Stochastic blockmodels: First steps,

    P. W. Holland, K. B. Laskey, and S. Leinhardt, “Stochastic blockmodels: First steps,”Social Networks, vol. 5, no. 2, pp. 109–137, 1983

  40. [48]

    Graph frequency analysis of brain signals,

    W. Huang, L. Goldsberry, N. F. Wymbs, S. T. Grafton, D. S. Bassett, and A. Ribeiro, “Graph frequency analysis of brain signals,”IEEE J. Sel. Topics Signal Process., vol. 10, no. 7, pp. 1189–1203, 2016

  41. [49]

    Decoupling of brain function from structure reveals regional behavioral specialization in humans,

    M. G. Preti and D. Van De Ville, “Decoupling of brain function from structure reveals regional behavioral specialization in humans,”Nat. Commun., vol. 10, pp. 1–7, 2019

  42. [50]

    Sigmoid-weighted linear units for neural network function approximation in reinforcement learning,

    S. Elfwing, E. Uchibe, and K. Doya, “Sigmoid-weighted linear units for neural network function approximation in reinforcement learning,” Neural Networks, vol. 107, pp. 3–11, 2018

  43. [51]

    A kernel two-sample test,

    A. Gretton, K. M. Borgwardt, M. J. Rasch, B. Sch ¨olkopf, and A. Smola, “A kernel two-sample test,”J. Mach. Learn. Res., vol. 13, no. 25, pp. 723–773, 2012

  44. [52]

    Lipschitz regularity of deep neural networks: Analysis and efficient estimation,

    K. Scaman and A. Virmaux, “Lipschitz regularity of deep neural networks: Analysis and efficient estimation,” inProc. Adv. Neural. Inf. Process. Syst., 2018, pp. 1–10

  45. [53]

    Learning by transference: Training graph neural networks on growing graphs,

    J. Cervi ˜no, L. Ruiz, and A. Ribeiro, “Learning by transference: Training graph neural networks on growing graphs,”IEEE Trans. Signal Process., vol. 71, pp. 233–247, 2023

Pith tools

Reviewed July 9, 2026 · model on record in the stance chip above.