pith. sign in

arxiv: 2509.07877 · v2 · pith:JAHBCO2Unew · submitted 2025-09-09 · 🧮 math.AP · math.OC· math.PR

Mean field control with absorption

Pierre Cardaliaguet , Joe Jackson , Panagiotis E. Souganidis This is my paper

Pith reviewed 2026-05-18 17:52 UTC · model grok-4.3

classification 🧮 math.AP math.OCmath.PR
keywords mean field controlabsorptionHamilton-Jacobi equationsub-probability measuresconvergence of particle systemsboundary conditionsinfinite-dimensional PDE
0
0 comments X

The pith

The N-particle value in mean field control with absorption converges to the limiting problem on sub-probability measures.

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

This paper examines a mean field control problem where particles disappear upon hitting the boundary of a smooth domain. It shows that the value function for systems with a finite number of particles approaches the value of a continuum limit as the particle count increases to infinity. The limit is characterized by a Hamilton-Jacobi equation posed on the space of sub-probability measures, which track the distribution of remaining particles with total mass at most one. Proving a comparison principle for this equation is central to establishing the convergence result. Readers interested in approximating large interacting systems with boundary effects, such as in fluid dynamics or population models, would find this convergence useful for simplifying computations.

Core claim

The value of the N-particle problem converges in a suitable sense towards the value of the limiting problem as N tends to infinity. This is supported by establishing a comparison principle for the Hamilton-Jacobi equation on the space of sub-probability measures on the smooth domain.

What carries the argument

A Hamilton-Jacobi equation defined on the space of sub-probability measures, coupled with boundary conditions from particle absorption.

If this is right

  • The comparison principle guarantees uniqueness of the solution to the infinite-dimensional Hamilton-Jacobi equation.
  • Convergence allows the limiting equation to serve as an approximation for large-scale particle control problems.
  • The use of sub-probability measures accounts for the loss of mass due to absorption at the boundary.
  • The hierarchy of equations for finite N provides a consistent approximation scheme.

Where Pith is reading between the lines

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

  • This framework could be tested in numerical experiments comparing finite-N simulations to the mean-field equation for small domains.
  • Extensions might include stochastic noise in the dynamics or different absorption rates.
  • Similar convergence results could apply to mean field games with absorption rather than control.
  • Applications in optimal transport or crowd dynamics with exits might benefit from this analysis.

Load-bearing premise

The domain must be smooth and the dynamics sufficiently regular so that the coupled hierarchy of Hamilton-Jacobi equations is well-posed.

What would settle it

Finding a specific smooth domain and dynamics where the N-particle values fail to converge to the mean-field value as N increases would falsify the claim.

read the original abstract

In this paper we study a mean field control problem in which particles are absorbed when they reach the boundary of a smooth domain. The value of the N-particle problem is described by a hierarchy of Hamilton-Jacobi equations which are coupled through their boundary conditions. The value function of the limiting problem; meanwhile, solves a Hamilton-Jacobi equation set on the space of sub-probability measures on the smooth domain, i.e. the space of non-negative measures with total mass at most one. Our main contributions are (i) to establish a comparison principle for this novel infinite-dimensional Hamilton-Jacobi equation and (ii) to prove that the value of the N-particle problem converges in a suitable sense towards the value of the limiting problem as N tends to infinity.

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 a mean field control problem with absorption at the boundary of a smooth domain. The N-particle value function satisfies a hierarchy of Hamilton-Jacobi equations coupled through boundary conditions that encode absorption (mass removal when particles hit the boundary, reducing to a lower-particle system). The mean-field limit is formulated as a Hamilton-Jacobi equation on the space of sub-probability measures. The main results are (i) a comparison principle for this infinite-dimensional equation and (ii) convergence of the N-particle values to the limiting value as N tends to infinity.

Significance. If rigorously established, the results would provide a technically nontrivial extension of mean-field control to open systems with absorption, using sub-probability measures to track mass loss. The comparison principle on the space of measures is a substantive contribution that could support further work on viscosity solutions in infinite dimensions. The convergence statement, if complete, would be the central applied payoff. The paper ships no machine-checked proofs or reproducible code, but the abstract claims independent proofs of comparison and convergence.

major comments (1)
  1. [§4] §4 (Convergence theorem): The comparison principle (Theorem 3.2) is used to identify the limit, but the argument for passing to the limit inside the coupled boundary conditions (which encode absorption and the resulting mass defect) appears to rely on abstract regularity of the domain and dynamics without quantitative estimates on hitting times or tightness in the sub-probability topology. This step is load-bearing for the claim that every limit point satisfies the limiting HJ equation; an explicit justification controlling the total-mass loss is needed to close the argument.
minor comments (2)
  1. [§2] Notation for the space of sub-probability measures is introduced without an explicit reference to the precise topology (e.g., weak vs. narrow) used for the viscosity solution; this should be stated once in §2.
  2. [§2.2] The hierarchy of boundary conditions for the N-particle problem is described in words; a short displayed equation summarizing the coupling (e.g., the absorption term when one coordinate hits ∂Ω) would improve readability.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their thorough review and valuable feedback on our work concerning mean field control with absorption. The comments help clarify the presentation of the convergence result. We respond to the major comment point by point below, and we plan to incorporate revisions to strengthen the justification for passing to the limit in the boundary conditions.

read point-by-point responses

  1. Referee: [§4] §4 (Convergence theorem): The comparison principle (Theorem 3.2) is used to identify the limit, but the argument for passing to the limit inside the coupled boundary conditions (which encode absorption and the resulting mass defect) appears to rely on abstract regularity of the domain and dynamics without quantitative estimates on hitting times or tightness in the sub-probability topology. This step is load-bearing for the claim that every limit point satisfies the limiting HJ equation; an explicit justification controlling the total-mass loss is needed to close the argument.

    Authors: We appreciate the referee's careful analysis of the convergence argument in Section 4. The proof proceeds by showing that the N-particle value functions are uniformly bounded and equicontinuous in a suitable topology on sub-probability measures, which implies tightness. The passage to the limit in the coupled boundary conditions is justified by the continuity of the absorption mechanism with respect to weak convergence of measures, combined with the fact that the total mass loss is controlled by the integral of the boundary flux, which is bounded due to the smoothness of the domain and the boundedness of the control set. We acknowledge that the current presentation relies on these properties without spelling out the quantitative bounds on hitting times. To make this step more transparent, we will add a new lemma providing explicit estimates on the expected mass loss up to time t, using Gronwall-type arguments and the regularity assumptions on the domain. This will ensure that every limit point indeed satisfies the limiting Hamilton-Jacobi equation with the correct mass defect. Thus, we will revise the manuscript accordingly. revision: yes

Circularity Check

0 steps flagged

No circularity: comparison principle and convergence established as independent results

full rationale

The paper's central claims rest on two distinct contributions: a comparison principle for the infinite-dimensional Hamilton-Jacobi equation on sub-probability measures, and a separate convergence argument showing that N-particle values approach the limiting value. These steps are derived from the well-posedness of the hierarchy of HJ equations with absorption boundary conditions and standard viscosity solution techniques, without any reduction of the limit to a fitted quantity, self-definition, or load-bearing self-citation chain. The derivation remains self-contained against external PDE benchmarks and does not rename known results or smuggle ansatzes via prior work.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Review performed from abstract only; ledger entries are inferred from stated setting rather than verified in the body.

axioms (1)
  • domain assumption The domain is smooth
    Required for the boundary conditions of the hierarchy of Hamilton-Jacobi equations.

pith-pipeline@v0.9.0 · 5655 in / 1123 out tokens · 33456 ms · 2026-05-18T17:52:58.643731+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.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Comparison of viscosity solutions for a class of non-linear PDEs on the space of finite nonnegative measures

    math.PR 2026-01 unverdicted novelty 7.0

    A comparison principle for viscosity solutions of nonlinear PDEs on finite nonnegative measures is proved and used to characterize the value function of a controlled branching McKean-Vlasov diffusion as the unique vis...

Reference graph

Works this paper leans on

40 extracted references · 40 canonical work pages · cited by 1 Pith paper · 1 internal anchor

  1. [1]

    Stable solutions in potential mean field game systems

    Ariela Briani and Pierre Cardaliaguet. Stable solutions in potential mean field game systems. NoDEA Nonlinear Differential Equations Appl. , 25(1):Paper No. 1, 26, 2018

    work page 2018

  2. [2]

    Mean field games with absorption and common noise with a model of bank run

    Matteo Burzoni and Luciano Campi. Mean field games with absorption and common noise with a model of bank run. Stochastic Processes and their Applications , 164:206--241, 2023

    work page 2023

  3. [3]

    Mean field control and finite agent approximation for regime-switching jump diffusions

    Erhan Bayraktar, Alekos Cecchin, and Prakash Chakraborty. Mean field control and finite agent approximation for regime-switching jump diffusions. Appl. Math. Optim. , 88(2), may 2023

    work page 2023

  4. [4]

    Viscosity solutions of fully second-order hjb equations in the wasserstein space

    Erhan Bayraktar, Hang Cheung, Ibrahim Ekren, Jinniao Qiu, Ho Man Tai, and Xin Zhang. Viscosity solutions of fully second-order hjb equations in the wasserstein space. 2025

    work page 2025

  5. [5]

    Large deviation properties of weakly interacting processes via weak convergence methods

    Amarjit Budhiraja, Paul Dupuis, and Markus Fischer. Large deviation properties of weakly interacting processes via weak convergence methods . The Annals of Probability , 40(1):74 -- 102, 2012

    work page 2012

  6. [6]

    Comparison for semi-continuous viscosity solutions for second order PDE s on the wasserstein space

    Erhan Bayraktar, Ibrahim Ekren, Xihao He, and Xin Zhang. Comparison for semi-continuous viscosity solutions for second order PDE s on the wasserstein space. 2025

    work page 2025

  7. [7]

    Stochastic optimal transport and hamilton-jacobi-bellman equations on the set of probability measures

    Charles Bertucci. Stochastic optimal transport and hamilton-jacobi-bellman equations on the set of probability measures. Preprint,arXiv:2306.04283 [math.AP], 2023

    work page arXiv 2023

  8. [8]

    Convergence rate of particle system for second-order pdes on wasserstein space

    Erhan Bayraktar, Ibrahim Ekren, and Xin Zhang. Convergence rate of particle system for second-order pdes on wasserstein space. 2024

    work page 2024

  9. [9]

    Comparison of viscosity solutions for a class of second-order pdes on the wasserstein space

    Erhan Bayraktar, Ibrahim Ekren, and Xin Zhang. Comparison of viscosity solutions for a class of second-order pdes on the wasserstein space. Communications in Partial Differential Equations , 50(4):570--613, 2025

    work page 2025

  10. [10]

    Souganidis

    Charles Bertucci, Pierre-Louis Lions, and Panagiotis E. Souganidis. Optimal control of the dyson equation and large deviations for hermitian random matrices. In preparation

    work page

  11. [11]

    Probabilistic Theory of Mean Field Games with Applications , volume 2 of Probability Theory and Stochastic Modelling

    Ren\' e Carmona and Fran c ois Delarue. Probabilistic Theory of Mean Field Games with Applications , volume 2 of Probability Theory and Stochastic Modelling . Springer International Publishing, 2018

    work page 2018

  12. [12]

    Probabilistic Theory of Mean Field Games with Applications , volume 1 of Probability Theory and Stochastic Modelling

    Ren\' e Carmona and Fran c ois Delarue. Probabilistic Theory of Mean Field Games with Applications , volume 1 of Probability Theory and Stochastic Modelling . Springer International Publishing, 2018

    work page 2018

  13. [13]

    Quantitative convergence for mean field control with common noise and degenerate idiosyncratic noise

    Alekos Cecchin, Samuel Daudin, Joe Jackson, and Mattia Martini. Quantitative convergence for mean field control with common noise and degenerate idiosyncratic noise. 2024

    work page 2024

  14. [14]

    Souganidis

    Pierre Cardaliaguet, Samuel Daudin, Joe Jackson, and Panagiotis E. Souganidis. An algebraic convergence rate for the optimal control of M c K ean- V lasov dynamics. SIAM J. Control Optim. , 61(6):3341--3369, 2023

    work page 2023

  15. [15]

    N -player games and mean-field games with absorption

    Luciano Campi and Markus Fischer. N -player games and mean-field games with absorption . The Annals of Applied Probability , 28(4):2188 -- 2242, 2018

    work page 2018

  16. [16]

    N -player games and mean-field games with smooth dependence on past absorptions

    Luciano Campi, Maddalena Ghio, and Giuilia Livieri. N -player games and mean-field games with smooth dependence on past absorptions . Annales de l’Institut Henri Poincaré - Probabilitates et Statistiques , 57(4):1901 -- 1939, 2021

    work page 1901

  17. [17]

    Souganidis

    Pierre Cardaliaguet, Joe Jackson, Nikiforos Mimikos-Stamatopoulos, and Panagiotis E. Souganidis. Sharp convergence rates for mean field control in the region of strong regularity. arXiv , 2312.11373, 2023

    work page arXiv 2023

  18. [18]

    Conforti, R.C

    G. Conforti, R.C. Kraaij, and D. Tonon. Hamilton–jacobi equations for controlled gradient flows: The comparison principle. Journal of Functional Analysis , 284(9):109853, 2023

    work page 2023

  19. [19]

    Kraaij, and Daniela Tonon

    Giovanni Conforti, Richard C. Kraaij, and Daniela Tonon. Hamilton--jacobi equations for controlled gradient flows: cylindrical test functions. 2 2023

    work page 2023

  20. [20]

    Kraaij, Luca Tamanini, and Daniela Tonon

    Giovanni Conforti, Richard C. Kraaij, Luca Tamanini, and Daniela Tonon. Hamilton--jacobi equations for wasserstein controlled gradient flows: existence of viscosity solutions. 1 2024

    work page 2024

  21. [21]

    On the optimal rate for the convergence problem in mean field control

    Samuel Daudin, François Delarue, and Joe Jackson. On the optimal rate for the convergence problem in mean field control. Journal of Functional Analysis , 287(12):110660, 2024

    work page 2024

  22. [22]

    Extended mean field control problem: a propagation of chaos result

    Mao Fabrice Djete. Extended mean field control problem: a propagation of chaos result . Electronic Journal of Probability , 27(none):1 -- 53, 2022

    work page 2022

  23. [23]

    Well-posedness of hamilton-jacobi equations in the wasserstein space: non-convex hamiltonians and common noise

    Samuel Daudin, Joe Jackson, and Benjamin Seeger. Well-posedness of hamilton-jacobi equations in the wasserstein space: non-convex hamiltonians and common noise. Communications in Partial Differential Equations , 50(1-2):1--52, 2025

    work page 2025

  24. [24]

    Mckean- V lasov optimal control: Limit theory and equivalence between different formulations

    Mao Fabrice Djete, Dylan Possamai, and Xiaolu Tan. Mckean- V lasov optimal control: Limit theory and equivalence between different formulations. Mathematics of Operations Research , 2022

    work page 2022

  25. [25]

    A comparison principle for semilinear hamilton-jacobi-bellman equations in the wasserstein space

    Samuel Daudin and Benjamin Seeger. A comparison principle for semilinear hamilton-jacobi-bellman equations in the wasserstein space. Calculus of Variations and Partial Differential Equations , 63(106), 2025

    work page 2025

  26. [26]

    Higher-order derivative estimates for the heat equation on a smooth domain

    Yoshinori Furoto and Tsukasa Iwabuchi. Higher-order derivative estimates for the heat equation on a smooth domain. 2025

    work page 2025

  27. [27]

    Finite dimensional approximations of H amilton- J acobi- B ellman equations in spaces of probability measures

    Wilfrid Gangbo, Sergio Mayorga, and Andrzej Swiech. Finite dimensional approximations of H amilton- J acobi- B ellman equations in spaces of probability measures. SIAM J. Math. Anal. , 53(2):1320--1356, 2021

    work page 2021

  28. [28]

    Jameson Graber and Ronnie Sircar

    P. Jameson Graber and Ronnie Sircar. Master equation for cournot mean field games of control with absorption. Journal of Differential Equations , 343:816--909, 2023

    work page 2023

  29. [29]

    Mean-field limit of particle systems with absorption

    Gaoyue Guo and Milica Tomasevic. Mean-field limit of particle systems with absorptio. arXiv:2310.10742 , 2024

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  30. [30]

    A stochastic McKean–Vlasov equation for absorbing diffusions on the half-line

    Ben Hambly and Sean Ledger. A stochastic McKean–Vlasov equation for absorbing diffusions on the half-line . The Annals of Applied Probability , 27(5):2698 -- 2752, 2017

    work page 2017

  31. [31]

    A McKean–Vlasov equation with positive feedback and blow-ups

    Ben Hambly, Sean Ledger, and Andreas S jmark. A McKean–Vlasov equation with positive feedback and blow-ups . The Annals of Applied Probability , 29(4):2338 -- 2373, 2019

    work page 2019

  32. [32]

    L imit theory for controlled M c K ean- V lasov dynamics

    Daniel Lacker. L imit theory for controlled M c K ean- V lasov dynamics. SIAM J. Control Optim. , 55:1641--1672, 2017

    work page 2017

  33. [33]

    P.-L. Lions. Cours au C oll\`ege de F rance, http://www.college-de-france. 2024

    work page 2024

  34. [34]

    At the mercy of the common noise: blow-ups in a conditional McKean–Vlasov Problem

    Sean Ledger and Andreas S jmark. At the mercy of the common noise: blow-ups in a conditional McKean–Vlasov Problem . Electronic Journal of Probability , 26(none):1 -- 39, 2021

    work page 2021

  35. [35]

    Finite dimensional approximations of hamilton–jacobi–bellman equations for stochastic particle systems with common noise

    Sergio Mayorga and Andrzej Swiech. Finite dimensional approximations of hamilton–jacobi–bellman equations for stochastic particle systems with common noise. SIAM Journal on Control and Optimization , 61(2):820--851, 2023

    work page 2023

  36. [36]

    Particle systems with singular interaction through hitting times: Application in systemic risk modeling

    Sergey Nadtochiy and Mykhaylo Shkolnikov. Particle systems with singular interaction through hitting times: Application in systemic risk modeling . The Annals of Applied Probability , 29(1):89 -- 129, 2019

    work page 2019

  37. [37]

    May 2020 mean field systems on networks, with singular interaction through hitting times

    Sergey Nadtochiy and Mykhaylo Shkolnikov. May 2020 mean field systems on networks, with singular interaction through hitting times. Annals of Probability , 48:1520--1556, 2020

    work page 2020

  38. [38]

    Mete Soner and Qinxin Yan

    H. Mete Soner and Qinxin Yan. Viscosity solutions for mckean–vlasov control on a torus. SIAM Journal on Control and Optimization , 62(2):903--923, 2024

    work page 2024

  39. [39]

    Mete Soner and Qinxin Yan

    H. Mete Soner and Qinxin Yan. Viscosity solutions of the eikonal equation on the wasserstein space. Applied Mathematics and Optimization , 90(1), 2024

    work page 2024

  40. [40]

    Viscosity solutions for HJB equations on the process space: Application to mean field control with common noise

    Jianjun Zhou, Nizar Touzi, and Jianfeng Zhang. Viscosity solutions for HJB equations on the process space: Application to mean field control with common noise. 2024

    work page 2024