pith. sign in

arxiv: 2507.03265 · v3 · submitted 2025-07-04 · 🧮 math.PR

Graphon particle systems with common noise

Erhan Bayraktar , Xihao He , Donghan Kim This is my paper

Pith reviewed 2026-05-19 06:48 UTC · model grok-4.3

classification 🧮 math.PR
keywords graphon particle systemscommon noiselaw of large numbersMcKean-Vlasov SDEgeneralized Wasserstein metricsweak convergencenon-Markovian processesempirical measures
0
0 comments X p. Extension

The pith

Graphon particle systems with common noise obey a law of large numbers for their empirical and interaction measures.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper examines systems of particles whose interactions are shaped by a graphon and whose motions are driven by both private idiosyncratic noise and a shared common noise. It proves that as the number of particles grows, the empirical distribution of particle states converges to a deterministic interaction measure. The argument relies on generalized Wasserstein metrics and weak-convergence arguments that accommodate the memory effects created by the common noise. A reader interested in large-scale networked models would care because many real systems, from financial markets to biological populations, experience simultaneous external shocks that ordinary mean-field limits do not capture.

Core claim

We study a nonlinear graphon particle system driven by both idiosyncratic and common noise, where interactions are governed by a graphon and represented as positive finite measures. Each particle evolves via a McKean-Vlasov-type SDE with graphon-weighted conditional laws. We prove a law of large numbers for the empirical and interaction measures, using generalized Wasserstein metrics and weak convergence techniques suited for the non-Markovian structure induced by common noise.

What carries the argument

Generalized Wasserstein metrics and weak convergence techniques adapted to non-Markovian dynamics induced by common noise

If this is right

  • The empirical measure converges to the interaction measure in the generalized Wasserstein metric.
  • The limiting behavior remains deterministic despite the presence of common noise.
  • The convergence result applies when interactions are represented as positive finite measures.
  • The techniques extend to systems whose dependence structure arises from shared randomness.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same metric framework could support efficient simulation of large networks subject to global shocks by replacing individual trajectories with the limiting measure.
  • Connections to mean-field games with common noise become feasible once the law of large numbers is available.
  • Similar arguments might apply to time-dependent graphons or to other sources of non-Markovianity beyond common noise.

Load-bearing premise

The non-Markovian structure induced by common noise can be handled using generalized Wasserstein metrics and weak convergence techniques.

What would settle it

Numerical simulation with increasing particle counts in which the generalized Wasserstein distance between the empirical measure and the limiting interaction measure does not approach zero.

read the original abstract

We study a nonlinear graphon particle system driven by both idiosyncratic and common noise, where interactions are governed by a graphon and represented as positive finite measures. Each particle evolves via a McKean-Vlasov-type SDE with graphon-weighted conditional laws. We prove a law of large numbers for the empirical and interaction measures, using generalized Wasserstein metrics and weak convergence techniques suited for the non-Markovian structure induced by common noise.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The manuscript studies nonlinear graphon particle systems driven by both idiosyncratic and common noise, with interactions encoded via graphons as positive finite measures. Particles obey McKean-Vlasov SDEs whose drift and diffusion coefficients depend on graphon-weighted conditional laws. The central claim is a law of large numbers for the empirical measure and the interaction measure, obtained by combining generalized Wasserstein metrics with weak-convergence arguments adapted to the non-Markovian dependence created by the shared Brownian motion.

Significance. If the limit theorem is established with the stated metric and convergence tools, the result supplies a rigorous mean-field approximation for large-scale graphon networks subject to global shocks. This is relevant to stochastic analysis of interacting particle systems on graphs and could support subsequent propagation-of-chaos or fluctuation analyses.

major comments (1)
  1. [Proof of the LLN (Section 3)] The abstract and the outline of the proof strategy assert that generalized Wasserstein metrics are 'suited' for the non-Markovian structure induced by common noise, yet the manuscript does not supply an explicit lemma verifying that the metric still yields tightness and uniform integrability when every particle is driven by the identical Brownian motion. Without such a control (e.g., a uniform bound on the second moments of the interaction measure under the common-noise filtration), the passage to the limit for the interaction measures remains unsecured.
minor comments (2)
  1. [Model setup] Notation for the graphon-weighted conditional law is introduced without an explicit display equation; adding a numbered display would improve readability when the same object reappears in the SDE and in the distance estimates.
  2. [Main result] The statement of the main theorem should list the precise integrability and Lipschitz assumptions on the coefficients and on the graphon; these are mentioned in passing but not collected in one place.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments on our manuscript. We address the major comment below and will incorporate the suggested clarification in the revised version.

read point-by-point responses

  1. Referee: [Proof of the LLN (Section 3)] The abstract and the outline of the proof strategy assert that generalized Wasserstein metrics are 'suited' for the non-Markovian structure induced by common noise, yet the manuscript does not supply an explicit lemma verifying that the metric still yields tightness and uniform integrability when every particle is driven by the identical Brownian motion. Without such a control (e.g., a uniform bound on the second moments of the interaction measure under the common-noise filtration), the passage to the limit for the interaction measures remains unsecured.

    Authors: We agree that isolating the tightness and uniform integrability controls would strengthen the presentation. The generalized Wasserstein metric is chosen precisely because it accommodates the shared Brownian motion by working with conditional laws under the common-noise filtration; the moment bounds follow from the linear growth assumptions on the coefficients and Gronwall-type estimates applied pathwise with respect to the common noise. In the revision we will add an explicit lemma (new Lemma 3.2) that states and proves the required uniform second-moment bound on the interaction measures and the resulting tightness in the generalized Wasserstein space, thereby making the passage to the limit fully rigorous and self-contained. revision: yes

Circularity Check

0 steps flagged

No circularity: standard stochastic analysis derivation

full rationale

This is a pure theoretical proof paper establishing a law of large numbers for empirical and interaction measures in a graphon-driven McKean-Vlasov system with common noise. The derivation relies on generalized Wasserstein metrics and weak convergence techniques applied to the non-Markovian structure. No parameters are fitted to data, no self-definitional loops appear in the equations, and no load-bearing steps reduce to self-citations or ansatzes imported from the authors' prior work. The central claims are derived from first-principles stochastic analysis without renaming known results or smuggling assumptions via citation chains. The paper is self-contained against external benchmarks in probability theory.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The claim rests on standard background results from stochastic analysis and graphon theory rather than new postulates or fitted quantities.

axioms (2)
  • domain assumption Existence and uniqueness of solutions to the McKean-Vlasov-type SDE with graphon-weighted conditional laws
    Required for the particle system to be well-defined before proving the law of large numbers.
  • domain assumption The graphon governs interactions represented as positive finite measures
    Standard modeling choice for dense graph limits in the particle system.

pith-pipeline@v0.9.0 · 5588 in / 1282 out tokens · 43098 ms · 2026-05-19T06:48:21.947954+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

32 extracted references · 32 canonical work pages

  1. [1]

    Aurell, A., Carmona, R., and Lauri` ere, M. (2022). Stochastic Graphon Games: II. The Linear-Quadratic Case. Applied Mathematics & Optimization , 85(3):39

    work page 2022

  2. [2]

    Bayraktar, E., Chakraborty, S., and Wu, R. (2023a). Graphon mean field systems. The Annals of Applied Probability, 33(5):3587 – 3619

    work page

  3. [3]

    and Kim, D

    Bayraktar, E. and Kim, D. (2024). Concentration of measure for graphon particle system. Advances in Applied Probability, 56(4):1279–1306

    work page 2024

  4. [4]

    and Wu, R

    Bayraktar, E. and Wu, R. (2022). Stationarity and uniform in time convergence for the graphon particle system. Stochastic Processes and their Applications , 150:532–568

    work page 2022

  5. [5]

    and Wu, R

    Bayraktar, E. and Wu, R. (2023). Graphon particle system: Uniform-in-time concentration bounds. Stochastic Processes and their Applications , 156:196–225

    work page 2023

  6. [6]

    Bayraktar, E., Wu, R., and Zhang, X. (2023b). Propagation of Chaos of Forward-Backward Stochastic Differential Equations with Graphon Interactions. Applied Mathematics & Opti- mization, 88

    work page

  7. [7]

    and Zhou, H

    Bayraktar, E. and Zhou, H. (2025+). Non-parametric estimates for graphon mean-field particle systems. To appear in Bernoulli. Available on ArXiv

    work page 2025

  8. [8]

    Billingsley, P. (2013). Convergence of probability measures. John Wiley & Sons

    work page 2013

  9. [9]

    Caines, P. E. and Huang, M. (2018). Graphon Mean Field Games and the GMFG Equations. In 2018 IEEE Conference on Decision and Control (CDC) , pages 4129–4134

    work page 2018

  10. [10]

    Caines, P. E. and Huang, M. (2021). Graphon mean field games and their equations. SIAM Journal on Control and Optimization , 59(6):4373–4399

    work page 2021

  11. [11]

    B., Graves, C

    Carmona, R., Cooney, D. B., Graves, C. V., and Lauri` ere, M. (2022). Stochastic graphon games: I. the static case. Mathematics of Operations Research, 47(1):750–778

    work page 2022

  12. [12]

    and Li, T

    Chen, Y. and Li, T. (2025). Graphon Particle Systems, Part II: Dynamics of Distributed Stochastic Continuum Optimization. arXiv:2407.02765

    work page arXiv 2025

  13. [13]

    Coppini, F. (2022). A note on Fokker-Planck equations and graphons. Journal of Statistical Physics, 187

    work page 2022

  14. [14]

    D., and Pham, H

    Coppini, F., Crescenzo, A. D., and Pham, H. (2025+). Nonlinear graphon mean-field systems. To appear in Stochastic Processes and Their Applications, available on ArXiv

    work page 2025

  15. [15]

    Mean field games: A toy model on an Erd¨ os-Renyi graph.ESAIM: Procs, 60:1–26

    Delarue, Fran¸ cois (2017). Mean field games: A toy model on an Erd¨ os-Renyi graph.ESAIM: Procs, 60:1–26

    work page 2017

  16. [16]

    F., Possama¨ ı, D., and Tan, X

    Djete, M. F., Possama¨ ı, D., and Tan, X. (2022). McKean–Vlasov optimal control: limit theory and equivalence between different formulations. Mathematics of Operations Research, 47(4):2891–2930

    work page 2022

  17. [17]

    Local-density dependent markov processes on graphons with epidemiological applications

    D´ aniel Keliger and Ill´ es Horv´ ath and B´ alint Tak´ acs (2022). Local-density dependent markov processes on graphons with epidemiological applications. Stochastic Processes and their Ap- plications, 148:324–352. 32

    work page 2022

  18. [18]

    and Guillin, A

    Fournier, N. and Guillin, A. (2015). On the rate of convergence in Wasserstein distance of the empirical measure. Probability theory and related fields, 162(3):707–738

    work page 2015

  19. [19]

    E., and Huang, M

    Gao, S., Caines, P. E., and Huang, M. (2021a). LQG graphon mean field games: Graphon invariant subspaces. In 2021 60th IEEE Conference on Decision and Control (CDC) , pages 5253–5260

    work page 2021

  20. [20]

    F., and Caines, P

    Gao, S., Tchuendom, R. F., and Caines, P. E. (2021b). Linear quadratic graphon field games. Communications in Information and Systems , 21:341–369

    work page

  21. [21]

    and Soret, A

    Lacker, D. and Soret, A. (2023). A label-state formulation of stochastic graphon games and approximate equilibria on large networks. Mathematics of Operations Research, 48(4):1987– 2018

    work page 2023

  22. [22]

    L., Liandrat, J., and Tournus, M

    Leblanc, H., Gouic, T. L., Liandrat, J., and Tournus, M. (2023). Extending the Wasserstein metric to positive measures. arXiv:2303.02183

    work page arXiv 2023

  23. [23]

    Lei, J. (2020). Convergence and concentration of empirical measures under Wasserstein distance in unbounded functional spaces. Bernoulli, 26(1):767 – 798

    work page 2020

  24. [24]

    Lov´ asz, L. (2012). Large networks and graph limits , volume 60. American Mathematical Soc

    work page 2012

  25. [25]

    and Ozdaglar, A

    Parise, F. and Ozdaglar, A. (2023). Graphon games: A statistical framework for network games and interventions. Econometrica, 91(1):191–225

    work page 2023

  26. [26]

    Sun, Y. (2006). The exact law of large numbers via Fubini extension and characterization of insurable risks. Journal of Economic Theory , 126(1):31–69

    work page 2006

  27. [27]

    Sznitman, A.-S. (1991). Topics in propagation of chaos. Ecole d’´ et´ e de probabilit´ es de Saint-Flour XIX—1989, 1464:165–251

    work page 1991

  28. [28]

    F., Caines, P

    Tchuendom, R. F., Caines, P. E., and Huang, M. (2020). On the master equation for linear quadratic graphon mean field games. In 2020 59th IEEE Conference on Decision and Control (CDC), pages 1026–1031

    work page 2020

  29. [29]

    F., Caines, P

    Tchuendom, R. F., Caines, P. E., and Huang, M. (2021). Critical nodes in graphon mean field games. In 2021 60th IEEE Conference on Decision and Control (CDC) , pages 166–170

    work page 2021

  30. [30]

    Vasal, D., Mishra, R., and Vishwanath, S. (2021). Sequential decomposition of graphon mean field games. In 2021 American Control Conference (ACC) , pages 730–736

    work page 2021

  31. [31]

    Villani, C. (2016). Optimal Transport: Old and New . Grundlehren der mathematischen Wissenschaften. Springer Berlin Heidelberg

    work page 2016

  32. [32]

    Villani, C. (2021). Topics in optimal transportation , volume 58. American Mathematical Soc. 33

    work page 2021