pith. sign in

arxiv: 2602.01138 · v2 · pith:Y2BY3HKFnew · submitted 2026-02-01 · 🧮 math.PR · math.AP

Rigorous derivation of the mean-field limit for the signal-dependent Keller-Segel system

Jinhuan Wang , Keyu Li , Hui Huang This is my paper

Pith reviewed 2026-05-21 14:49 UTC · model grok-4.3

classification 🧮 math.PR math.AP
keywords Keller-Segel systemmean-field limitpropagation of chaosinteracting particlessignal-dependent sensitivitystochastic particle modelalgebraic scaling
0
0 comments X

The pith

A stochastic interacting particle system converges in probability to the signal-dependent Keller-Segel equations under algebraic scaling.

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

The paper establishes that a system of stochastic particles with signal-dependent interactions converges to a deterministic two-dimensional Keller-Segel type PDE in the mean-field limit. It achieves this convergence in probability using stopping times to handle close encounters between particles, which allows an algebraic scaling regime rather than the weaker logarithmic one used before. From there it applies the relative-entropy method to prove strong L1 propagation of chaos with an algebraic rate of convergence. A sympathetic reader would care because this justifies deriving macroscopic chemotaxis models from microscopic rules with better quantitative control, relevant to biological aggregation phenomena.

Core claim

We rigorously derive a two-dimensional Keller-Segel type system with signal-dependent sensitivity from a stochastic interacting particle model. By employing suitably defined stopping times, we prove that the convergence of the interacting particle system towards the corresponding mean-field limit equations holds in probability under an algebraic scaling regime which improves upon existing results with logarithmic scaling. Building on this, we apply the relative-entropy method to obtain strong L1 propagation of chaos, and establish an algebraic convergence rate.

What carries the argument

Suitably defined stopping times that bound the minimal distance between particles, ensuring the interaction kernel stays integrable at the improved algebraic scaling.

If this is right

  • The interacting particle system converges in probability to the mean-field Keller-Segel equations.
  • Strong L1 propagation of chaos holds for the particle system.
  • An algebraic rate of convergence is obtained.
  • The derivation improves previous results by replacing logarithmic scaling with algebraic scaling.

Where Pith is reading between the lines

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

  • Similar stopping-time techniques could extend to other singular interaction kernels in aggregation models.
  • Algebraic rates might enable better error bounds in numerical simulations of chemotaxis.
  • Connections to other mean-field limits in kinetic theory could be explored using the same relative-entropy approach.

Load-bearing premise

The stopping times can be defined so that they control the minimal particle distance and keep the singular kernel integrable throughout the algebraic scaling regime.

What would settle it

A direct computation or simulation showing that without the stopping time cutoff, the probability of particles getting closer than the algebraic scale goes to one before the mean-field limit is reached.

read the original abstract

We rigorously derive a two-dimensional Keller-Segel type system with signal-dependent sensitivity from a stochastic interacting particle model. By employing suitably defined stopping times, we prove that the convergence of the interacting particle system towards the corresponding mean-field limit equations in probability under an algebraic scaling regime which improves upon existing results with logarithmic scaling. Building on this, we apply the relative-entropy method to obtain strong $L^1$ propagation of chaos, and establish an algebraic convergence rate.

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

3 major / 3 minor

Summary. The paper claims to rigorously derive the mean-field limit for a two-dimensional signal-dependent Keller-Segel system from a stochastic interacting particle model. Using suitably defined stopping times to control minimal inter-particle distances, the authors establish convergence in probability of the empirical measure to the mean-field PDE under an algebraic scaling regime in N, improving on prior logarithmic scaling results. They then apply the relative-entropy method to obtain strong L¹ propagation of chaos with an algebraic convergence rate.

Significance. If the stopping-time construction successfully delivers algebraic rates without hidden logarithmic factors from the 2D singularity, the result would constitute a technical advance in quantitative mean-field limits for singular kernels. The direct probabilistic argument combined with relative-entropy propagation of chaos provides both convergence in probability and strong quantitative estimates, which are useful for analyzing aggregation models in mathematical biology. The algebraic improvement, if verified, strengthens the applicability of such derivations beyond logarithmic regimes.

major comments (3)
  1. [§3.2] §3.2, definition of stopping time τ_N: The claim that τ_N keeps min_{i≠j} |X^i_t - X^j_t| ≳ N^{-β} with high probability up to algebraic order must be checked against the 2D Coulomb-type singularity; if the bound on P(τ_N < T) is obtained via a Gronwall estimate that produces only a logarithmic factor, the algebraic scaling improvement over existing logarithmic results would not hold.
  2. [§4] §4, probability estimate for particle-system convergence: The integrability of the interaction kernel up to the algebraic scaling is invoked after introducing the stopping times; the quantitative dependence of this estimate on the exit probability P(τ_N < T) needs explicit tracking to confirm that no logarithmic deterioration occurs.
  3. [§5] §5, relative-entropy propagation: The L¹ convergence rate obtained via relative entropy relies on the particle system remaining well-defined and controlled by τ_N up to time T with high probability; the entropy dissipation inequality should display the explicit algebraic rate after accounting for the stopping-time threshold.
minor comments (3)
  1. [Abstract] The abstract states the algebraic scaling but does not specify the admissible range for the exponent β; adding this detail would improve clarity.
  2. [Notation] Notation for the signal-dependent sensitivity function varies slightly between the particle system and the mean-field limit; consistent use of a single symbol would help readability.
  3. [Introduction] A few references to prior logarithmic-scaling works in the introduction could be expanded with one-sentence summaries of their rates to better highlight the improvement.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments on our manuscript. The points raised concern the technical details of the stopping-time construction and the preservation of algebraic rates, which we address below with clarifications and commitments to revise the presentation for greater explicitness.

read point-by-point responses

  1. Referee: [§3.2] §3.2, definition of stopping time τ_N: The claim that τ_N keeps min_{i≠j} |X^i_t - X^j_t| ≳ N^{-β} with high probability up to algebraic order must be checked against the 2D Coulomb-type singularity; if the bound on P(τ_N < T) is obtained via a Gronwall estimate that produces only a logarithmic factor, the algebraic scaling improvement over existing logarithmic results would not hold.

    Authors: We thank the referee for this important observation. The stopping time τ_N is constructed to halt when any pair reaches distance N^{-β} with β chosen sufficiently small. The bound on P(τ_N < T) is obtained via a direct probabilistic estimate: we control the expected number of near-collisions using the non-degenerate Brownian motion and the uniform bound on the drifts (which remain controlled up to the stopping time), combined with a union bound over the N(N-1)/2 pairs. This yields P(τ_N < T) ≤ C_T N^{-γ} for an explicit positive γ depending on β, without invoking Gronwall on the minimal-distance process itself. Consequently no logarithmic factor appears. We will add a dedicated remark after Lemma 3.3 clarifying this estimation strategy and the resulting algebraic decay. revision: yes

  2. Referee: [§4] §4, probability estimate for particle-system convergence: The integrability of the interaction kernel up to the algebraic scaling is invoked after introducing the stopping times; the quantitative dependence of this estimate on the exit probability P(τ_N < T) needs explicit tracking to confirm that no logarithmic deterioration occurs.

    Authors: We agree that explicit dependence tracking is necessary to confirm the algebraic rate. In the proof of convergence in probability, the error is split into the contribution on {τ_N ≥ T}, where the kernel is integrable down to distance N^{-β} and the algebraic scaling in N is preserved by standard moment bounds, and the contribution on {τ_N < T}, which is at most P(τ_N < T) times a uniform bound on the total variation of the empirical measures. Because our estimate for P(τ_N < T) is algebraic, the overall convergence rate remains algebraic. We will insert an explicit inequality displaying this decomposition and the dependence on P(τ_N < T) in the revised proof of Theorem 4.1. revision: yes

  3. Referee: [§5] §5, relative-entropy propagation: The L¹ convergence rate obtained via relative entropy relies on the particle system remaining well-defined and controlled by τ_N up to time T with high probability; the entropy dissipation inequality should display the explicit algebraic rate after accounting for the stopping-time threshold.

    Authors: We appreciate the referee's request for explicit display of the rate. The relative-entropy dissipation inequality is derived for the stopped processes; on the event {τ_N ≥ T} the minimal separation guarantees that the singular interaction term remains controlled, yielding the algebraic decay in the relative entropy. The contribution of the complementary event is bounded by P(τ_N < T) times a constant (from the a-priori L^1 bound on the densities), which is again algebraic. We will revise the statement of the propagation-of-chaos result and the corresponding proof in Section 5 to include the modified dissipation inequality that incorporates the stopping-time threshold and explicitly exhibits the algebraic rate. revision: yes

Circularity Check

0 steps flagged

No circularity: direct probabilistic convergence proof

full rationale

The derivation proceeds via a standard probabilistic argument: suitably chosen stopping times control minimal inter-particle distances to ensure integrability of the singular kernel up to algebraic scaling, followed by convergence in probability and relative-entropy propagation of chaos. No parameter is fitted to data and then relabeled as a prediction, no self-citation supplies a load-bearing uniqueness theorem or ansatz, and the stopping-time estimates are derived from the particle dynamics rather than presupposing the target mean-field limit. The argument is therefore self-contained against external benchmarks and does not reduce any claimed result to its own inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The paper relies on standard stochastic calculus and relative-entropy techniques; no free parameters or new entities are introduced in the abstract.

axioms (1)
  • standard math Standard Itô calculus and martingale estimates for the particle system
    Invoked to control the stochastic differential equations before applying stopping times.

pith-pipeline@v0.9.0 · 5598 in / 1094 out tokens · 40173 ms · 2026-05-21T14:49:29.206843+00:00 · methodology

discussion (0)

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

Forward citations

Cited by 1 Pith paper

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

  1. Quantitative propagation of chaos for particle systems with bounded kernels and multiplicative noise

    math.AP 2026-04 unverdicted novelty 6.0

    Quantitative propagation of chaos holds for particle systems with bounded drift kernels and multiplicative noise via an extension of the Jabin-Wang relative entropy framework using dynamic combinatorial analysis.

Reference graph

Works this paper leans on

41 extracted references · 41 canonical work pages · cited by 1 Pith paper

  1. [1]

    On mean field limits for dynamical systems.Journal of Statistical Physics, 164(1):1– 16, 2016

    Niklas Boers and Peter Pickl. On mean field limits for dynamical systems.Journal of Statistical Physics, 164(1):1– 16, 2016

    work page 2016

  2. [2]

    Two-dimensional signal-dependent parabolic-elliptic Keller-Segel system and its mean-field derivation.Journal of Differential Equations, 450:113712, 2026

    Lukas Bol, Li Chen, and Yue Li. Two-dimensional signal-dependent parabolic-elliptic Keller-Segel system and its mean-field derivation.Journal of Differential Equations, 450:113712, 2026

    work page 2026

  3. [3]

    Stochastic mean-field limit: non-lipschitz forces and swarm- ing.Mathematical Models and Methods in Applied Sciences, 21(11):2179–2210, 2011

    Fran¸ cois Bolley, Jos´ e A Canizo, and Jos´ e A Carrillo. Stochastic mean-field limit: non-lipschitz forces and swarm- ing.Mathematical Models and Methods in Applied Sciences, 21(11):2179–2210, 2011

    work page 2011

  4. [4]

    Mean field limit and quantitative estimates with singular attractive kernels.Duke Mathematical Journal, 172(13):2591–2641, 2023

    Didier Bresch, Pierre-Emmanuel Jabin, and Zhenfu Wang. Mean field limit and quantitative estimates with singular attractive kernels.Duke Mathematical Journal, 172(13):2591–2641, 2023

    work page 2023

  5. [5]

    Delayed blow-up for chemotaxis models with local sensing.Journal of the London Mathematical Society, 103(4):1596–1617, 2021

    Martin Burger, Philippe Lauren¸ cot, and Ariane Trescases. Delayed blow-up for chemotaxis models with local sensing.Journal of the London Mathematical Society, 103(4):1596–1617, 2021

    work page 2021

  6. [6]

    Propagation of chaos for the Vlasov–Poisson–Fokker–Planck equation with a polynomial cut-off.Communications in Contemporary Mathematics, 21(04):1850039, 2019

    Jos´ e A Carrillo, Young-Pil Choi, and Samir Salem. Propagation of chaos for the Vlasov–Poisson–Fokker–Planck equation with a polynomial cut-off.Communications in Contemporary Mathematics, 21(04):1850039, 2019

    work page 2019

  7. [7]

    Interacting particle approximation of cross-diffusion systems.arXiv preprint arXiv:2402.05094, 2024

    Jose Antonio Carrillo and Shuchen Guo. Interacting particle approximation of cross-diffusion systems.arXiv preprint arXiv:2402.05094, 2024

    work page arXiv 2024

  8. [8]

    Rigorous derivation of population cross- diffusion systems from moderately interacting particle systems.Journal of Nonlinear Science, 31(6):94, 2021

    Li Chen, Esther S Daus, Alexandra Holzinger, and Ansgar J¨ ungel. Rigorous derivation of population cross- diffusion systems from moderately interacting particle systems.Journal of Nonlinear Science, 31(6):94, 2021

    work page 2021

  9. [9]

    Rigorous mean-field limit and cross-diffusion.Zeitschrift f¨ ur ange- wandte Mathematik und Physik, 70(4):122, 2019

    Li Chen, Esther S Daus, and Ansgar J¨ ungel. Rigorous mean-field limit and cross-diffusion.Zeitschrift f¨ ur ange- wandte Mathematik und Physik, 70(4):122, 2019

    work page 2019

  10. [10]

    Mean field limit and propagation of chaos for a pedestrian flow model

    Li Chen, Simone G¨ ottlich, and Qitao Yin. Mean field limit and propagation of chaos for a pedestrian flow model. Journal of Statistical Physics, 166(2):211–229, 2017

    work page 2017

  11. [11]

    Quantitative convergence in relative entropy for a moderately interacting particle system onR d.Electronic Journal of Probability, 30:1–24, 2025

    Li Chen, Alexandra Holzinger, and Xiaokai Huo. Quantitative convergence in relative entropy for a moderately interacting particle system onR d.Electronic Journal of Probability, 30:1–24, 2025

    work page 2025

  12. [12]

    Fluctuations around the mean-field limit for attractive Riesz potentials in the moderate regime.arXiv preprint arXiv:2405.15128, 2024

    Li Chen, Alexandra Holzinger, and Ansgar J¨ ungel. Fluctuations around the mean-field limit for attractive Riesz potentials in the moderate regime.arXiv preprint arXiv:2405.15128, 2024

    work page arXiv 2024

  13. [13]

    On the mean-field limit of Vlasov-Poisson-Fokker-Planck equations.arXiv preprint arXiv:2505.13038, 2025

    Li Chen, Jinwook Jung, Peter Pickl, and Zhenfu Wang. On the mean-field limit of Vlasov-Poisson-Fokker-Planck equations.arXiv preprint arXiv:2505.13038, 2025

    work page arXiv 2025

  14. [14]

    Combined mean field limit and non-relativistic limit of Vlasov– Maxwell particle system to Vlasov–Poisson system.Journal of Mathematical Physics, 61(6), 2020

    Li Chen, Xin Li, Peter Pickl, and Qitao Yin. Combined mean field limit and non-relativistic limit of Vlasov– Maxwell particle system to Vlasov–Poisson system.Journal of Mathematical Physics, 61(6), 2020

    work page 2020

  15. [15]

    American Mathematical Society, 2022

    Lawrence C Evans.Partial differential equations, volume 19. American Mathematical Society, 2022

    work page 2022

  16. [16]

    Stripe formation in bacterial systems with density-suppressed motility.Physical review letters, 108(19):198102, 2012

    Xiongfei Fu, Lei-Han Tang, Chenli Liu, Jian-Dong Huang, Terence Hwa, and Peter Lenz. Stripe formation in bacterial systems with density-suppressed motility.Physical review letters, 108(19):198102, 2012

    work page 2012

  17. [17]

    Boundedness of classical solutions to a degenerate Keller–Segel type model with signal-dependent motilities.Acta Applicandae Mathematicae, 176:1–36, 2021

    Kentaro Fujie and Jie Jiang. Boundedness of classical solutions to a degenerate Keller–Segel type model with signal-dependent motilities.Acta Applicandae Mathematicae, 176:1–36, 2021

    work page 2021

  18. [18]

    Global existence for a kinetic model of pattern formation with density-suppressed motilities.Journal of Differential Equations, 269(6):5338–5378, 2020

    Kentarou Fujie and Jie Jiang. Global existence for a kinetic model of pattern formation with density-suppressed motilities.Journal of Differential Equations, 269(6):5338–5378, 2020

    work page 2020

  19. [19]

    Comparison methods for a Keller–Segel-type model of pattern formations with density-suppressed motilities.Calculus of Variations and Partial Differential Equations, 60(3):92, 2021

    Kentarou Fujie and Jie Jiang. Comparison methods for a Keller–Segel-type model of pattern formations with density-suppressed motilities.Calculus of Variations and Partial Differential Equations, 60(3):92, 2021

    work page 2021

  20. [20]

    Error estimate of a random particle blob method for the Keller-Segel equation

    Hui Huang and Jian-Guo Liu. Error estimate of a random particle blob method for the Keller-Segel equation. Mathematics of Computation, 86(308):2719–2744, 2017. CONVERGENCE IN PROBABILITY AND RELATIVE ENTROPY 19

    work page 2017

  21. [21]

    On the mean-field limit for the Vlasov–Poisson–Fokker–Planck system

    Hui Huang, Jian-Guo Liu, and Peter Pickl. On the mean-field limit for the Vlasov–Poisson–Fokker–Planck system. Journal of Statistical Physics, 181(5):1915–1965, 2020

    work page 1915

  22. [22]

    Mean field limit for stochastic particle systems.Active Particles, Volume 1: Advances in Theory, Models, and Applications, pages 379–402, 2017

    Pierre-Emmanuel Jabin and Zhenfu Wang. Mean field limit for stochastic particle systems.Active Particles, Volume 1: Advances in Theory, Models, and Applications, pages 379–402, 2017

    work page 2017

  23. [23]

    Quantitative estimates of propagation of chaos for stochastic systems with ˙W −1,∞ kernels.Inventiones Mathematicae, 214(1):523–591, 2018

    Pierre-Emmanuel Jabin and Zhenfu Wang. Quantitative estimates of propagation of chaos for stochastic systems with ˙W −1,∞ kernels.Inventiones Mathematicae, 214(1):523–591, 2018

    work page 2018

  24. [24]

    Critical mass on the Keller-Segel system with signal-dependent motility.Pro- ceedings of the American Mathematical Society, 148(11):4855–4873, 2020

    Hai-Yang Jin and Zhi-An Wang. Critical mass on the Keller-Segel system with signal-dependent motility.Pro- ceedings of the American Mathematical Society, 148(11):4855–4873, 2020

    work page 2020

  25. [25]

    A mean field limit for the Vlasov–Poisson system.Archive for Rational Mechanics and Analysis, 225(3):1201–1231, 2017

    Dustin Lazarovici and Peter Pickl. A mean field limit for the Vlasov–Poisson system.Archive for Rational Mechanics and Analysis, 225(3):1201–1231, 2017

    work page 2017

  26. [26]

    Optimal mass on the parabolic-elliptic-ODE minimal chemotaxis-haptotaxis inR 2

    Siying Li and Jinhuan Wang. Optimal mass on the parabolic-elliptic-ODE minimal chemotaxis-haptotaxis inR 2. Physica Scripta, 98(9):095223, 2023

    work page 2023

  27. [27]

    Convergence towards the population cross-diffusion system from stochastic many-particle system.Commun

    Yue Li, Li Chen, and Zhipeng Zhang. Convergence towards the population cross-diffusion system from stochastic many-particle system.Commun. Math. Res., 40:43–63, 2024

    work page 2024

  28. [28]

    American Mathematical Soc., 2001

    Elliott H Lieb and Michael Loss.Analysis, volume 14. American Mathematical Soc., 2001

    work page 2001

  29. [29]

    Genealogies and increasing propagation of chaos for Feynman-Kac and genetic models.The Annals of Applied Probability, 11(4):1166–1198, 2001

    Laurent Miclo and Pierre Del Moral. Genealogies and increasing propagation of chaos for Feynman-Kac and genetic models.The Annals of Applied Probability, 11(4):1166–1198, 2001

    work page 2001

  30. [30]

    Springer, 2006

    David Nualart.The Malliavin calculus and related topics. Springer, 2006

    work page 2006

  31. [31]

    A law of large numbers for moderately interacting diffusion processes.Zeitschrift f¨ ur Wahrscheinlichkeitstheorie und verwandte Gebiete, 69(2):279–322, 1985

    Karl Oelschl¨ ager. A law of large numbers for moderately interacting diffusion processes.Zeitschrift f¨ ur Wahrscheinlichkeitstheorie und verwandte Gebiete, 69(2):279–322, 1985

    work page 1985

  32. [32]

    A fluctuation theorem for moderately interacting diffusion processes.Probability theory and related fields, 74(4):591–616, 1987

    Karl Oelschl¨ ager. A fluctuation theorem for moderately interacting diffusion processes.Probability theory and related fields, 74(4):591–616, 1987

    work page 1987

  33. [33]

    Karl Oelschl¨ ager. On the derivation of reaction-diffusion equations as limit dynamics of systems of moderately interacting stochastic processes.Probability Theory and Related Fields, 82(4):565–586, 1989

    work page 1989

  34. [34]

    Quantitative approximation of the Keller-Segel and Burgers equations by moderately interacting particles.arXiv preprint arXiv:2004.03177, 2020

    C Olivera, A Richard, and M Tomaˇ sevic. Quantitative approximation of the Keller-Segel and Burgers equations by moderately interacting particles.arXiv preprint arXiv:2004.03177, 2020

    work page arXiv 2004

  35. [35]

    Quantitative particle approximation of nonlinear Fokker-Planck equations with singular kernel, 2021

    Christian Olivera, Alexandre Richard, and Milica Tomasevic. Quantitative particle approximation of nonlinear Fokker-Planck equations with singular kernel, 2021

    work page 2021

  36. [36]

    Quantitative particle approximation of nonlinear Fokker-Planck equations with singular kernel.Journal of Statistical Physics, 24(2):691–749, 2023

    Christian Olivera, Alexandre Richard, and Milica Tomasevic. Quantitative particle approximation of nonlinear Fokker-Planck equations with singular kernel.Journal of Statistical Physics, 24(2):691–749, 2023

    work page 2023

  37. [37]

    Mean field limit for Coulomb-type flows.Duke mathematical journal, 169:2887–2935, 2020

    Sylvia Serfaty and Mitia Duerinckx. Mean field limit for Coulomb-type flows.Duke mathematical journal, 169:2887–2935, 2020

    work page 2020

  38. [38]

    The derivation of chemotaxis equations as limit dynamics of moderately interacting stochastic many-particle systems.SIAM Journal on Applied Mathematics, 61(1):183–212, 2000

    Angela Stevens. The derivation of chemotaxis equations as limit dynamics of moderately interacting stochastic many-particle systems.SIAM Journal on Applied Mathematics, 61(1):183–212, 2000

    work page 2000

  39. [39]

    Topics in propagation of chaos

    Alain-Sol Sznitman. Topics in propagation of chaos. InEcole d’´ et´ e de probabilit´ es de Saint-Flour XIX—1989, pages 165–251. Springer, 2006

    work page 1989

  40. [40]

    Effects of signal-dependent motilities in a Keller–Segel-type reaction–diffusion system.Mathematical Models and Methods in Applied Sciences, 27(09):1645–1683, 2017

    Youshan Tao and Michael Winkler. Effects of signal-dependent motilities in a Keller–Segel-type reaction–diffusion system.Mathematical Models and Methods in Applied Sciences, 27(09):1645–1683, 2017

    work page 2017

  41. [41]

    A review of mathematical topics in collisional kinetic theory.Handbook of mathematical fluid dynamics, 1:71–74, 2002

    C´ edric Villani. A review of mathematical topics in collisional kinetic theory.Handbook of mathematical fluid dynamics, 1:71–74, 2002

    work page 2002