pith. sign in

arxiv: 2604.17639 · v1 · submitted 2026-04-19 · 🧮 math.AP · math.OC

Convergence of Potential Mean-Field Games via Lyapunov Methods

Felix H\"ofer This is my paper

Pith reviewed 2026-05-10 05:17 UTC · model grok-4.3

classification 🧮 math.AP math.OC
keywords mean-field gamespotential gamesLyapunov functionalconvergencestationary equilibriumKuramoto modelinfinite horizontorus domain
0
0 comments X

The pith

In potential mean-field games, every weak limit of a time-dependent equilibrium is a stationary equilibrium.

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

The paper proves that for discounted infinite-horizon potential mean-field games on the d-dimensional torus, every weak limit point of a time-dependent equilibrium as time tends to infinity is a stationary equilibrium. This holds without any monotonicity assumptions on the costs. The proof relies on constructing a novel Lyapunov functional that decreases along solutions of the time-dependent system. Consequently, if there is a unique stationary equilibrium, then the time-dependent equilibria converge to it. The authors also provide a new uniqueness criterion for stationary equilibria and demonstrate the result for the subcritical Kuramoto mean-field game, where all equilibria converge to the incoherent solution.

Core claim

We consider discounted infinite-horizon potential mean-field games on the d-dimensional torus. Without imposing monotonicity assumptions, we prove that every weak limit point of a time-dependent equilibrium, as time tends to infinity, is a stationary equilibrium. As a consequence, equilibria converge whenever the stationary solution is unique. The short proof is based on a novel Lyapunov functional for the time-dependent MFG system. We also provide a new uniqueness criterion for stationary equilibria. Finally, we apply our results to the subcritical Kuramoto MFG, showing that every equilibrium converges to the incoherent solution.

What carries the argument

A novel Lyapunov functional for the time-dependent MFG system, built from the potential functional that defines the game costs.

If this is right

  • Equilibria converge whenever the stationary equilibrium is unique.
  • A new criterion for uniqueness of stationary equilibria becomes available.
  • All equilibria in the subcritical Kuramoto MFG converge to the incoherent solution.
  • Long-time behavior of non-monotone potential MFGs can be analyzed without monotonicity assumptions.

Where Pith is reading between the lines

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

  • The Lyapunov construction may extend to other potential game structures outside mean-field settings.
  • Numerical simulations of Kuramoto-type models could test convergence speed and stability under perturbations.
  • Similar functionals might help study stability in potential MFGs on non-torus domains if boundary conditions permit.
  • The uniqueness criterion could be checked against existing monotone cases to see if it recovers known results.

Load-bearing premise

The mean-field game must derive its costs from a single potential functional and be posed as a discounted infinite-horizon problem on the d-dimensional torus.

What would settle it

A counterexample would be a potential MFG on the torus where some time-dependent equilibrium has a weak limit point as t tends to infinity that fails to satisfy the stationary MFG equations.

read the original abstract

We consider discounted infinite-horizon potential mean-field games (MFGs) on the $d$-dimensional torus. Without imposing monotonicity assumptions, we prove that every weak limit point of a time-dependent equilibrium, as time tends to infinity, is a stationary equilibrium. As a consequence, equilibria converge whenever the stationary solution is unique. The short proof is based on a novel Lyapunov functional for the time-dependent MFG system. We also provide a new uniqueness criterion for stationary equilibria. Finally, we apply our results to the subcritical Kuramoto MFG studied by Carmona, Cormier, and Soner, showing that every equilibrium converges to the incoherent solution.

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 considers discounted infinite-horizon potential mean-field games on the d-dimensional torus. Without monotonicity assumptions, it proves that every weak limit point of a time-dependent equilibrium as time tends to infinity is a stationary equilibrium, via a novel Lyapunov functional constructed from the potential structure. It also supplies a new uniqueness criterion for stationary equilibria and applies the results to the subcritical Kuramoto MFG, concluding that every equilibrium converges to the incoherent solution.

Significance. If the central argument holds, the result is significant because it establishes long-time convergence for potential MFGs by dropping the standard monotonicity hypothesis and instead using a Lyapunov functional that exploits the potential structure in the discounted torus setting. The self-contained nature of the proof and the explicit application to the Kuramoto model (showing convergence to the incoherent state) are strengths. The approach provides a clean route when monotonicity is unavailable and yields falsifiable predictions for specific models.

minor comments (3)
  1. §2 (setup): the precise definition of the potential functional and the associated Hamiltonian should be recalled explicitly before the Lyapunov construction to improve readability for readers unfamiliar with the Carmona–Cormier–Soner setting.
  2. §3 (Lyapunov functional): the passage from the time-dependent system to the derivative of the functional along trajectories is clear in outline but would benefit from one additional line justifying the integration-by-parts step under the weak topology.
  3. §5 (Kuramoto application): the statement that the incoherent solution is the unique stationary equilibrium should be cross-referenced to the new uniqueness criterion in §4 rather than left as a citation.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive summary of our work on discounted infinite-horizon potential mean-field games and for recommending minor revision. The recognition of the novel Lyapunov functional and its application to the subcritical Kuramoto model is appreciated. Since the report lists no major comments, we have prepared revisions addressing any minor points and provide the following point-by-point responses.

Circularity Check

0 steps flagged

No significant circularity; derivation self-contained via Lyapunov construction

full rationale

The paper's central result—that weak limit points of time-dependent equilibria converge to stationary equilibria—rests on a Lyapunov functional explicitly constructed from the potential structure of the MFG, the discount factor, and the torus geometry. This construction is independent of the target convergence statement and does not reduce by definition or fitting to the result itself. No self-citation chains, ansatz smuggling, or renaming of known results appear in the load-bearing steps; the uniqueness criterion is stated as new, and the Kuramoto application is to an external model. The proof is therefore internally consistent and non-circular.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The result rests on the potential structure of the MFG (costs come from a single functional) and standard properties of the torus; no free parameters or invented entities are introduced.

axioms (2)
  • domain assumption The mean-field game is potential, so individual costs derive from a common functional whose derivative recovers the running cost and terminal cost.
    Invoked in the setup of potential MFGs and used to construct the Lyapunov functional.
  • standard math The state space is the d-dimensional torus with periodic boundary conditions.
    Allows weak convergence arguments without boundary terms.

pith-pipeline@v0.9.0 · 5394 in / 1298 out tokens · 31294 ms · 2026-05-10T05:17:18.350649+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

27 extracted references · 27 canonical work pages

  1. [1]

    Achdou, F

    Y. Achdou, F. J. Buera, J.-M. Lasry, P.-L. Lions, and B. Moll. Partial differential equation models in macroeco- nomics.Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 372 (2028):20130397, 2014

    work page 2028

  2. [2]

    Achdou, J

    Y. Achdou, J. Han, J.-M. Lasry, P.-L. Lions, and B. Moll. Income and wealth distribution in macroeconomics: A continuous-time approach.The review of economic studies, 89(1):45–86, 2022

    work page 2022

  3. [3]

    Bakry, I

    D. Bakry, I. Gentil, M. Ledoux, et al.Analysis and geometry of Markov diffusion operators, volume 103. Springer, 2014

    work page 2014

  4. [4]

    Bardi and E

    M. Bardi and E. Feleqi. Nonlinear elliptic systems and mean-field games.Nonlinear Differential Equations and Applications NoDEA, 23(4):44, 2016

    work page 2016

  5. [5]

    Başar, B

    T. Başar, B. Djehiche, and H. Tembine.Mean-Field-Type Game Theory I-II. Springer Nature, 2026

    work page 2026

  6. [6]

    Bayraktar and X

    E. Bayraktar and X. Zhang. Solvability of infinite horizon McKean–Vlasov FBSDEs in mean field control problems and games.Applied Mathematics & Optimization, 87(1):13, 2023

    work page 2023

  7. [7]

    Cardaliaguet and M

    P. Cardaliaguet and M. Masoero. Weak KAM theory for potential MFG.Journal of Differential Equations, 268 (7):3255–3298, 2020

    work page 2020

  8. [8]

    Cardaliaguet and A

    P. Cardaliaguet and A. Porretta. Long time behavior of the master equation in mean field game theory.Analysis & PDE, 12(6):1397–1453, 2019

    work page 2019

  9. [9]

    Cardaliaguet, J.-M

    P. Cardaliaguet, J.-M. Lasry, P.-L. Lions, and A. Porretta. Long time average of mean field games.Networks & Heterogeneous Media, 7(2), 2012

    work page 2012

  10. [10]

    Cardaliaguet, J.-M

    P. Cardaliaguet, J.-M. Lasry, P.-L. Lions, and A. Porretta. Long time average of mean field games with a nonlocal coupling.SIAM Journal on Control and Optimization, 51(5):3558–3591, 2013

    work page 2013

  11. [11]

    Carmona, F

    R. Carmona, F. Delarue, et al.Probabilistic theory of mean field games with applications I-II. Springer, 2018

    work page 2018

  12. [12]

    Carmona, Q

    R. Carmona, Q. Cormier, and H. M. Soner. Synchronization in a Kuramoto mean field game.Communications in Partial Differential Equations, 48(9):1214–1244, 2023

    work page 2023

  13. [13]

    Carmona, L

    R. Carmona, L. Tangpi, and K. Zhang. A probabilistic approach to discounted infinite horizon and invariant mean field games.arXiv preprint arXiv:2407.03642, 2024

    work page arXiv 2024

  14. [14]

    Carmona, Q

    R. Carmona, Q. Cormier, and H. M. Soner. Kuramoto mean field game with intrinsic frequencies.arXiv preprint arXiv:2509.18000, 2025

    work page arXiv 2025

  15. [15]

    Cesaroni and M

    A. Cesaroni and M. Cirant. Stationary equilibria and their stability in a Kuramoto MFG with strong interaction. Communications in Partial Differential Equations, 49(1-2):121–147, 2024

    work page 2024

  16. [16]

    M. Cirant. On the existence of oscillating solutions in non-monotone mean-field games.Journal of Differential Equations, 266(12):8067–8093, 2019. 20 Höfer Convergence of Potential MFGs

    work page 2019

  17. [17]

    Cirant and L

    M. Cirant and L. Nurbekyan. The variational structure and time-periodic solutions for mean-field games systems. arXiv preprint arXiv:1804.08943, 2018

    work page arXiv 2018

  18. [18]

    Cirant and A

    M. Cirant and A. Porretta. Long time behavior and turnpike solutions in mildly non-monotone mean field games. ESAIM: Control, Optimisation and Calculus of Variations, 27:86, 2021

    work page 2021

  19. [19]

    Claisse, G

    J. Claisse, G. Conforti, Z. Ren, and S. Wang. Mean field optimization problem regularized by fisher information. The Annals of Applied Probability, 36(2):1002–1047, 2026

    work page 2026

  20. [20]

    Émery and J

    M. Émery and J. E. Yukich. A simple proof of the logarithmic Sobolev inequality on the circle.Séminaire de probabilités, 21:173–175, 1987

    work page 1987

  21. [21]

    D. A. Gomes, J. Mohr, and R. R. Souza. Discrete time, finite state space mean field games.Journal de mathématiques pures et appliquées, 93(3):308–328, 2010

    work page 2010

  22. [22]

    Höfer and H

    F. Höfer and H. M. Soner. Optimal control and potential games in the mean field.arXiv preprint arXiv:2408.00733, 2024

    work page arXiv 2024

  23. [23]

    Höfer and H

    F. Höfer and H. M. Soner. Synchronization games.Mathematics of Operations Research, 2025

    work page 2025

  24. [24]

    Huang, P

    M. Huang, P. E. Caines, and R. P. Malhamé. Individual and mass behaviour in large population stochastic wireless power control problems: Centralized and Nash equilibrium solutions. In42nd IEEE International Conference on Decision and Control, volume 1, pages 98–103, 2003

    work page 2003

  25. [25]

    Lasry and P.-L

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

    work page 2007

  26. [26]

    H. Yin, P. G. Mehta, S. P. Meyn, and U. V. Shanbhag. Bifurcation analysis of a heterogeneous mean-field oscillator game model. In2011 50th IEEE Conference on Decision and Control and European Control Conference, pages 3895–3900. IEEE, 2011

    work page 2011

  27. [27]

    H. Yin, P. G. Mehta, S. P. Meyn, and U. V. Shanbhag. Synchronization of coupled oscillators is a game.IEEE Transactions on Automatic Control, 57(4):920–935, 2011. 21

    work page 2011