pith. sign in

arxiv: 2605.14901 · v1 · pith:P6YBDFOEnew · submitted 2026-05-14 · 🧮 math.OC · math.PR

Non--exchangeable mean field games with moderate interactions and common noise

Mao Fabrice Djete This is my paper

Pith reviewed 2026-06-30 19:58 UTC · model grok-4.3

classification 🧮 math.OC math.PR
keywords mean field gamesnon-exchangeable populationsgraphon interactionscommon noiserelaxed equilibriaNash equilibriamoderate interactions
0
0 comments X

The pith

Limit points of approximate closed-loop Nash equilibria from non-exchangeable finite-player games converge to relaxed solutions of the limiting mean field game, with averaged payoffs converging, and conversely every relaxed equilibrium admi

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

The paper examines mean field games in large populations whose agents are not exchangeable, using a graphon to encode heterogeneous large-scale interactions together with a rescaled local kernel for moderate density-dependent effects and common noise. It formulates a relaxed version of the limiting game indexed by a label u in [0,1] and proves existence of solutions under continuity and non-degeneracy conditions on the coefficients. The central result establishes a two-way asymptotic link: every limit point of approximate closed-loop Nash equilibria of the finite-player system is a relaxed equilibrium of the mean field game, and every such relaxed equilibrium can be realized approximately by Markovian strategies in the finite system. This supplies a complete characterization of equilibrium behavior in the non-exchangeable setting.

Core claim

Every limit point of approximate closed-loop Nash equilibria of the finite-player system is a relaxed solution of the limiting mean field game, and the corresponding averaged equilibrium payoffs converge. Conversely, every relaxed mean field game equilibrium can be approximated by Markovian approximate Nash equilibria of the finite-player systems.

What carries the argument

Relaxed formulation of the non-exchangeable mean field game in which the representative player indexed by u in [0,1] interacts through a graphon-weighted local density and a graphon-induced environment law, adapted to common noise.

If this is right

  • Existence of relaxed equilibria holds under general continuity and non-degeneracy assumptions on the coefficients.
  • Under additional convexity assumptions, relaxed equilibria can be realized in strict form.
  • In the deterministic case without common noise, deterministic equilibria exist and strict equilibria admit a nonlinear Feynman-Kac probabilistic characterization.
  • The bidirectional convergence supplies a complete asymptotic characterization of equilibrium behavior for non-exchangeable games with moderate interactions and common noise.

Where Pith is reading between the lines

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

  • The graphon-plus-kernel structure separates global heterogeneity from local density effects, allowing the same limit theory to cover both network-like and spatially distributed agent systems.
  • Markovian finite-player approximations indicate that history-independent controls suffice once the continuum limit is taken.
  • The relaxed formulation absorbs the common noise while preserving the ability to pass to the limit, suggesting the same machinery could handle other shared random factors such as aggregate shocks.

Load-bearing premise

The continuity and non-degeneracy assumptions on the coefficients, together with the specific graphon-type structure and rescaled local kernel, are needed to establish existence of relaxed equilibria and the convergence between finite and limiting systems.

What would settle it

A concrete sequence of finite-player approximate closed-loop Nash equilibria whose limit points fail to satisfy the relaxed mean field game equations, or a relaxed equilibrium for which no sequence of Markovian finite-player strategies achieves the required approximation.

read the original abstract

We study mean field games for large non--exchangeable populations with moderate local interactions and common noise. The finite--player system is driven by two complementary interaction mechanisms : a graphon--type structure, which encodes heterogeneous large--scale interactions between agents, and a rescaled local kernel, which produces a density-dependent interaction term in the limit. The limiting model is a non--exchangeable mean field game in which the representative player is indexed by a label \(u\in[0,1]\), interacts through a graphon--weighted local density, and is affected by a graphon--induced environment law. We introduce a relaxed formulation of the limiting mean field game, adapted to the presence of common noise, and prove existence under general continuity and non--degeneracy assumptions. Under additional convexity assumptions, relaxed equilibria can be realized in strict form. In the deterministic case without common noise, we obtain deterministic equilibria and provide a probabilistic characterization of strict equilibria through a nonlinear Feynman--Kac representation. We then establish the asymptotic connection with the finite--player game. We prove that every limit point of approximate closed--loop Nash equilibria is a relaxed solution of the limiting mean field game, and that the corresponding averaged equilibrium payoffs converge. Conversely, every relaxed mean field game equilibrium can be approximated by Markovian approximate Nash equilibria of the finite--player systems. These results give a complete asymptotic characterization of equilibrium behavior for non--exchangeable games with moderate interactions and 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

0 major / 3 minor

Summary. The manuscript develops a theory of non-exchangeable mean field games with moderate local interactions (via graphon-type structures and rescaled local kernels) and common noise. It introduces a relaxed formulation of the limiting MFG, proves existence of equilibria under continuity and non-degeneracy assumptions (with strict realization under added convexity), provides a nonlinear Feynman-Kac characterization in the deterministic case, and establishes a two-way asymptotic link: every limit point of approximate closed-loop Nash equilibria of the finite-player system is a relaxed MFG equilibrium (with convergence of averaged payoffs), and conversely every relaxed MFG equilibrium can be approximated by Markovian approximate Nash equilibria of the finite-player systems.

Significance. If the derivations hold, the work supplies a complete asymptotic characterization of equilibrium behavior for non-exchangeable populations with heterogeneous large-scale and local-density interactions plus common noise. The bidirectional convergence result (limit points to equilibria and approximability in the converse direction) together with the relaxed formulation adapted to common noise constitute a clear technical advance over prior exchangeable or non-local MFG frameworks. The deterministic Feynman-Kac representation is a standard but cleanly applied tool that strengthens the strict-equilibrium case.

minor comments (3)
  1. The abstract invokes 'graphon-weighted local density' and 'graphon-induced environment law' without a one-sentence gloss; a brief parenthetical would improve immediate readability for readers outside the immediate subfield.
  2. Notation for the label space u ∈ [0,1] and the precise form of the rescaled kernel could be introduced with a short display equation already in the introduction to avoid forward references.
  3. The statement that 'relaxed equilibria can be realized in strict form' under convexity would benefit from an explicit pointer to the relevant theorem number once the sections are finalized.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive evaluation, detailed summary of our contributions, and recommendation of minor revision. No specific major comments were raised in the report.

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper establishes existence of relaxed equilibria and asymptotic convergence between finite-player Nash equilibria and the mean-field limit through direct arguments in stochastic analysis. These rely on explicit continuity/non-degeneracy assumptions for compactness, the graphon-weighted interaction structure, and standard tools such as nonlinear Feynman-Kac representations. No step reduces by construction to a fitted input, self-defined quantity, or load-bearing self-citation; the two-way characterization is derived from the problem formulation and assumptions without circular reduction. The derivation is therefore self-contained against external mathematical benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claims rest on standard domain assumptions from stochastic control rather than new free parameters or invented entities.

axioms (1)
  • domain assumption Continuity and non-degeneracy assumptions on the model coefficients and graphon
    Invoked to guarantee existence of relaxed equilibria and convergence results.

pith-pipeline@v0.9.1-grok · 5790 in / 1250 out tokens · 33021 ms · 2026-06-30T19:58:57.869819+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

42 extracted references · 26 canonical work pages · 2 internal anchors

  1. [1]

    Aliprantis and K

    C. Aliprantis and K. Border.Infinite dimensional analysis: a hitchhiker’s guide. Springer–Verlag Berlin Heidelberg, 3rd edition, 2006

    2006

  2. [2]

    D. G. Aronson and J. Serrin. Local behavior of solutions of quasilinear parabolic equations.Archive for Rational Mechanics and Analysis volume, 25:81–122, 1967. doi: 10.1137/S0363012996313549. URLhttps://doi.org/10. 1007/BF00281291

    work page doi:10.1137/s0363012996313549 1967

  3. [3]

    Aurell, R

    A. Aurell, R. A. Carmona, and M. Laurière. Stochastic graphon games: II. the linear-quadratic case.Applied Mathematics & Optimization, 85, 2021. URLhttps://api.semanticscholar.org/CorpusID:235195871. 46

    2021

  4. [4]

    Bayraktar, S

    E. Bayraktar, S. Chakraborty, and R. Wu. Graphon mean field systems.The Annals of Applied Probability, 33(5): 3587 – 3619, 2023. doi: 10.1214/22-AAP1901. URLhttps://doi.org/10.1214/22-AAP1901

    work page doi:10.1214/22-aap1901 2023

  5. [5]

    Bayraktar, R

    E. Bayraktar, R. Wu, and X. Zhang. Propagation of chaos of forward–backward stochastic differential equations with graphon interactions.Appl. Math. Optim., 88(1), 2023. ISSN 0095-4616. doi: 10.1007/s00245-023-09996-y. URLhttps://doi.org/10.1007/s00245-023-09996-y

    work page doi:10.1007/s00245-023-09996-y 2023

  6. [6]

    Graphon particle systems with common noise

    E. Bayraktar, X. He, and D. Kim. Graphon particle systems with common noise. 2026. URLhttps://arxiv. org/abs/2507.03265

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  7. [7]

    P. E. Caines and M. Huang. Graphon mean field games and their equations.SIAM J. Control. Optim., 59: 4373–4399, 2020

    2020

  8. [8]

    Cao and M

    Z. Cao and M. Laurière. Probabilistic analysis of graphon mean field control. 2025. URLhttps://arxiv.org/ abs/2505.19664

    work page arXiv 2025

  9. [9]

    Cardaliaguet

    P. Cardaliaguet. The convergence problem in mean field games with local coupling.Applied Mathematics & Optimization, 76, 2017. doi: 10.1007/s00245-017-9434-0. URLhttps://doi.org/10.1007/s00245-017-9434-0

    work page doi:10.1007/s00245-017-9434-0 2017

  10. [10]

    Carmona and F

    R. Carmona and F. Delarue.Probabilistic theory of mean field games with applications I, volume 83 ofProbability theory and stochastic modelling. Springer International Publishing, 2018

    2018

  11. [11]

    Carmona and F

    R. Carmona and F. Delarue.Probabilistic theory of mean field games with applications II, volume 84 ofProbability theory and stochastic modelling. Springer International Publishing, 2018

    2018

  12. [12]

    Carmona, F

    R. Carmona, F. Delarue, and D. Lacker. Mean field games with common noise.The Annals of Probability, 44(6): 3740 – 3803, 2016. doi: 10.1214/15-AOP1060. URLhttps://doi.org/10.1214/15-AOP1060

    work page doi:10.1214/15-aop1060 2016

  13. [13]

    Coppini, A

    F. Coppini, A. D. Crescenzo, and H. Pham. Nonlinear graphon mean-field systems.Stochastic Processes and their Applications, 2024. URLhttps://api.semanticscholar.org/CorpusID:267636698

    2024

  14. [14]

    A. D. Crescenzo, M. Fuhrman, I. Kharroubi, and H. Pham. Mean-field control of non exchangeable systems. 2024. URLhttps://arxiv.org/abs/2407.18635

    work page arXiv 2024

  15. [15]

    Crucianelli and L

    C. Crucianelli and L. Tangpi. Interacting particle systems on sparse W-random graphs. 2024. URLhttps: //arxiv.org/abs/2410.11240

    work page arXiv 2024

  16. [16]

    M. F. Djete. Large population games with interactions through controls and common noise: convergence results and equivalence between open-loop and closed-loop controls.ESAIM: COCV, 29:39, 2023. doi: 10.1051/cocv/2023005. URLhttps://doi.org/10.1051/cocv/2023005

    work page doi:10.1051/cocv/2023005 2023

  17. [17]

    M. F. Djete. Stackelberg mean field games: convergence and existence results to the problem of principal with multiple agents in competition. 2023. URLhttps://arxiv.org/abs/2309.00640

    work page arXiv 2023

  18. [18]

    M. F. Djete. A non-exchangeable mean field control problem with controlled interactions. 2025. URLhttps: //arxiv.org/abs/2511.00288

    work page arXiv 2025

  19. [19]

    M. F. Djete and N. Touzi. Mean field game of mutual holding.The Annals of Applied Probability, 34(6):4999 – 5031, 2024. doi: 10.1214/23-AAP1993. URLhttps://doi.org/10.1214/23-AAP1993

    work page doi:10.1214/23-aap1993 2024

  20. [20]

    M. F. Djete, D. Possamaï, and X. Tan. Mckean–vlasov optimal control: Limit theory and equivalence between different formulations.Mathematics of Operations Research, 0(0):null, 0. doi: 10.1287/moor.2021.1232. URL https://doi.org/10.1287/moor.2021.1232

    work page doi:10.1287/moor.2021.1232 2021

  21. [21]

    M. F. Djete, D. Possamaï, and X. Tan. McKean–Vlasov optimal control: The dynamic programming principle. The Annals of Probability, 50(2):791 – 833, 2022. doi: 10.1214/21-AOP1548. URLhttps://doi.org/10.1214/ 21-AOP1548

    work page doi:10.1214/21-aop1548 2022

  22. [22]

    Capacities, Measurable Selection and Dynamic Programming Part I: Abstract Framework

    N. El Karoui and X. Tan. Capacities, measurable selection and dynamic programming part I: abstract framework. arXiv preprint arXiv:1310.3363, 2013

    work page internal anchor Pith review Pith/arXiv arXiv 2013

  23. [23]

    El Karoui, S

    N. El Karoui, S. Peng, and M.-C. Quenez. Backward stochastic differential equations in finance.Mathematical Finance, 7(1):1–71, 1997. 47

    1997

  24. [24]

    Flandoli, M

    F. Flandoli, M. Ghio, and G. Livieri. N-player games and mean field games of moderate interactions.Applied Mathematics & Optimization, 85, 2022. doi: 10.1007/s00245-022-09834-7. URLhttps://doi.org/10.1007/ s00245-022-09834-7

    work page doi:10.1007/s00245-022-09834-7 2022

  25. [25]

    S. Gao, R. F. Tchuendom, and P. E. Caines. Linear quadratic graphon field games.ArXiv, abs/2006.03964, 2020. URLhttps://api.semanticscholar.org/CorpusID:219530959

    work page arXiv 2006

  26. [26]

    Huang, R

    M. Huang, R. Malhamé, and P. Caines. Large population stochastic dynamic games: closed–loop McKean–Vlasov systems and the Nash certainty equivalence principle.Communications in Information & Systems, 6(3):221–252, 2006

    2006

  27. [27]

    Huang, P

    M. Huang, P. Caines, and R. Malhamé. The Nash certainty equivalence principle and McKean–Vlasov systems: an invariance principle and entry adaptation. In D. Castanon and J. Spall, editors,46th IEEE conference on decision and control, 2007, pages 121–126. IEEE, 2007

    2007

  28. [28]

    Jabin, D

    P.-E. Jabin, D. Poyato, and J. Soler. Mean-field limit of non-exchangeable systems. 2024. URLhttps://arxiv. org/abs/2112.15406

    work page arXiv 2024

  29. [29]

    Jourdain and S

    B. Jourdain and S. Méléard. Propagation of chaos and fluctuations for a moderate model with smooth initial data. Annales de l’institut Henri Poincaré, Probabilités et Statistiques(B), 34(6):727–766, 1998

    1998

  30. [30]

    Kharroubi, S

    I. Kharroubi, S. Mekkaoui, and H. Pham. Stochastic maximum principle for optimal control problem of non exchangeable mean field systems. 2025. URLhttps://arxiv.org/abs/2506.05595

    work page arXiv 2025

  31. [31]

    Krylov.Controlled diffusion processes, volume 14 ofStochastic modelling and applied probability

    N. Krylov.Controlled diffusion processes, volume 14 ofStochastic modelling and applied probability. Springer– Verlag New York, 1980

    1980

  32. [32]

    Kurtz, T

    G. Kurtz, T. Weak and strong solutions of general stochastic models.Electronic Communications in Probability, 19(58):1–16, 2014

    2014

  33. [33]

    D. Lacker. On the convergence of closed-loop Nash equilibria to the mean field game limit.The Annals of Applied Probability, 30(4):1693 – 1761, 2020. doi: 10.1214/19-AAP1541. URLhttps://doi.org/10.1214/19-AAP1541

    work page doi:10.1214/19-aap1541 2020

  34. [34]

    Lacker and L

    D. Lacker and L. L. Flem. Closed-loop convergence for mean field games with common noise.The Annals of Applied Probability, 33(4):2681 – 2733, 2023. doi: 10.1214/22-AAP1876. URLhttps://doi.org/10.1214/22-AAP1876

    work page doi:10.1214/22-aap1876 2023

  35. [35]

    Lacker and A

    D. Lacker and A. Soret. A label-state formulation of stochastic graphon games and approximate equilibria on large networks.Mathematics of Operations Research, 48(4):1987–2018, 2023. doi: 10.1287/moor.2022.1329. URL https://doi.org/10.1287/moor.2022.1329

    work page doi:10.1287/moor.2022.1329 1987

  36. [36]

    Lacker, M

    D. Lacker, M. Shkolnikov, and J. Zhang. Superposition and mimicking theorems for conditional Mckean-Vlasov equations.arXiv preprint arXiv:2004.00099, 2020

    work page arXiv 2004

  37. [37]

    Lasry and P.-L

    J.-M. Lasry and P.-L. Lions. Mean field games.Japanese Journal of Mathematics, 2(1):229–260, 2007

    2007

  38. [38]

    L. M. Lovász. Large networks and graph limits. InColloquium Publications, 2012. URLhttps://api. semanticscholar.org/CorpusID:27057044

    2012

  39. [39]

    Ma and J

    J. Ma and J. Zhang. Representation theorems for backward stochastic differential equations.The Annals of Applied Probability, 12(4):1390 – 1418, 2002. doi: 10.1214/aoap/1037125868. URLhttps://doi.org/10.1214/ aoap/1037125868

    work page doi:10.1214/aoap/1037125868 2002

  40. [40]

    Méléard and S

    S. Méléard and S. Roelly-Coppoletta. A propagation of chaos result for a system of particles with moderate interaction.Stochastic Processes and their Applications, 26:317–332, 1987

    1987

  41. [41]

    Oelschläger

    K. Oelschläger. A law of large numbers for moderately interacting diffusion processes.Zeitschrift für Wahrschein- lichkeitstheorie und verwandte Gebiete, 69(2):279–322, 1985

    1985

  42. [42]

    Tangpi and X

    L. Tangpi and X. Zhou. Optimal investment in a large population of competitive and heterogeneous agents.Finance and Stochastics, 28(2):497–551, April 2024. doi: 10.1007/s00780-023-00527-9. URLhttps://ideas.repec.org/ a/spr/finsto/v28y2024i2d10.1007_s00780-023-00527-9.html. 48

    work page doi:10.1007/s00780-023-00527-9 2024