pith. machine review for the scientific record. sign in

arxiv: 2604.23462 · v1 · submitted 2026-04-25 · 🧮 math.PR

Recognition: unknown

Invariant measures for the open KPZ equation: the Gaussian case

James Bona-Landry
Authors on Pith no claims yet

Pith reviewed 2026-05-08 07:15 UTC · model grok-4.3

classification 🧮 math.PR
keywords open KPZ equationinvariant measuresBrownian motionstationary measuresGaussian noiseboundary conditionsStein's equation
0
0 comments X

The pith

The open KPZ equation with constant boundary slopes admits drifted Brownian motion as an invariant measure modulo height shifts.

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

The paper proves that a specific random function remains statistically the same after evolving according to the open KPZ equation. It focuses on Gaussian noise and boundary conditions where the spatial derivative of the height is fixed to a constant value at both ends of an interval. Using Stein's equation, it shows that Brownian motion with matching drift is stationary, except for overall height shifts. This gives an explicit description of the long-time statistics for the interface in the open setting.

Core claim

The paper considers the open KPZ equation with boundary conditions requiring the derivative of the height function to equal alpha at both boundaries. It shows that a Brownian motion with constant drift alpha is invariant for this dynamics, up to adding a constant to the height function. The proof uses Stein's equation and integration by parts to verify that the generator applied to suitable test functions has zero expectation under this measure.

What carries the argument

Stein's equation combined with integration by parts applied to the generator of the open KPZ dynamics under fixed-slope Neumann boundary conditions.

If this is right

  • The law of the height function initialized as drifted Brownian motion stays unchanged over time except for a possible uniform vertical shift.
  • The invariance holds for every real value of the boundary slope parameter.
  • The same measure is stationary for the linear Edwards-Wilkinson equation as a special case.

Where Pith is reading between the lines

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

  • The technique might extend to non-Gaussian noise if the integration-by-parts identities can be adapted.
  • This stationary measure could serve as a starting point for analyzing fluctuations around equilibrium in open systems.
  • Results for the open interval may help classify stationary states when boundaries have different slopes at each end.

Load-bearing premise

The boundary conditions must allow Stein's equation and integration by parts to proceed without producing extra terms that break the invariance, and the equation must be well-posed enough that the drifted Brownian motion is a stationary measure.

What would settle it

A long-time numerical simulation of the open KPZ equation started from a drifted Brownian path, checking whether the distribution of the evolved height function matches the initial law up to a uniform height shift.

read the original abstract

In [arXiv:2409.08465], Quastel and Gu use Stein's equation and integration by parts to give a direct proof that drifted Brownian motions are stationary (modulo height shifts) for the full-line KPZ equation. In this article, we consider the open KPZ equation with boundary conditions $\partial_x h(t,0) = \partial_x h(t,1) = \alpha$ for a general real parameter $\alpha$, and emulate the approach of Quastel and Gu to provide a similar proof that Brownian motion with constant drift $\alpha$ is invariant (modulo height shifts) in this case.

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 emulates the Stein-equation and integration-by-parts argument of Quastel-Gu (arXiv:2409.08465) to prove that Brownian motion with constant drift α is invariant, modulo height shifts, for the open KPZ equation on [0,1] subject to Neumann boundary conditions ∂_x h(t,0)=∂_x h(t,1)=α.

Significance. If the boundary cancellations are verified, the result supplies a direct, non-circular verification of the stationary measure for the open KPZ equation, extending the full-line case without introducing fitted parameters or additional assumptions on the noise. This is useful for the analysis of KPZ dynamics in bounded domains with prescribed slope boundaries.

major comments (1)
  1. [proof of the main theorem] The integration-by-parts step in the proof of the main invariance statement: the generator of the open KPZ contains a Laplacian term whose integration against a test functional F produces boundary evaluations of the form F'(∂_x h - α) evaluated at x=0 and x=1. The manuscript must exhibit an explicit cancellation of these terms when the measure is drifted Brownian motion; without this calculation the invariance identity does not follow from the full-line argument.
minor comments (2)
  1. State the precise function space in which the test functionals F are taken and confirm that they are compatible with the Neumann boundary conditions.
  2. Add a short remark comparing the open-boundary generator to the full-line generator used by Quastel-Gu, highlighting exactly where the new boundary terms appear.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their careful reading and for identifying the need to make the boundary cancellations explicit in the integration-by-parts argument. We address this point below and will revise the manuscript to strengthen the proof.

read point-by-point responses

  1. Referee: The integration-by-parts step in the proof of the main invariance statement: the generator of the open KPZ contains a Laplacian term whose integration against a test functional F produces boundary evaluations of the form F'(∂_x h - α) evaluated at x=0 and x=1. The manuscript must exhibit an explicit cancellation of these terms when the measure is drifted Brownian motion; without this calculation the invariance identity does not follow from the full-line argument.

    Authors: We agree with the referee that an explicit verification of the boundary terms is required for the invariance identity to hold. While the manuscript emulates the Stein-equation and integration-by-parts strategy of Quastel-Gu, the current write-up does not spell out the cancellation of the Laplacian boundary contributions in sufficient detail. We will revise the proof of the main theorem by adding a self-contained computation of the integration by parts for the Laplacian term. Under the law of Brownian motion with drift α, the Neumann boundary conditions ensure that the evaluations involving (∂_x h - α) at x=0 and x=1 cancel (in the sense appropriate to the test functional F), so that the boundary terms vanish identically. This addition will render the argument independent of the full-line case and complete the proof. revision: yes

Circularity Check

0 steps flagged

No circularity: direct emulation of independent external proof

full rationale

The paper's central derivation explicitly emulates the Stein-equation plus integration-by-parts argument from the independent Quastel-Gu work (arXiv:2409.08465) and applies it to the open interval with Neumann boundary conditions. No self-citations appear, no parameters are fitted to data then renamed as predictions, and no definitions or uniqueness claims reduce to the paper's own outputs. The abstract and description present the adaptation as a straightforward extension whose validity rests on verifying boundary-term cancellation under the drifted-Brownian measure; this verification, if performed, would be independent content rather than a tautology. The result is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the assumption that the probabilistic technique transfers directly from the infinite-line to the open setting without new obstructions from the boundaries.

axioms (1)
  • domain assumption The open KPZ equation is well-posed under the given Neumann boundary conditions for Gaussian noise.
    Invoked implicitly to apply Stein's method and integration by parts to establish invariance.

pith-pipeline@v0.9.0 · 5394 in / 1367 out tokens · 62538 ms · 2026-05-08T07:15:17.675627+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

12 extracted references · 11 canonical work pages

  1. [1]

    Steady state of the KPZ equation on an interval and Liouville quantum mechanics

    Guillaume Barraquand and Pierre Le Doussal. “Steady state of the KPZ equation on an interval and Liouville quantum mechanics”. In:Europhysics Letters137.6 (Mar. 2022), p. 61003.issn: 1286- 4854.doi:10.1209/0295-5075/ac25a9.url:http://dx.doi.org/10.1209/0295-5075/ac25a9

    work page doi:10.1209/0295-5075/ac25a9.url:http://dx.doi.org/10.1209/0295-5075/ac25a9 2022

  2. [2]

    Markov processes related to the stationary measure for the open KPZ equa- tion

    Włodek Bryc et al. “Markov processes related to the stationary measure for the open KPZ equa- tion”. In:Probability Theory and Related Fields185.1–2 (Feb. 2022), pp. 353–389.issn: 1432-2064. doi:10.1007/s00440-022-01110-7.url:http://dx.doi.org/10.1007/s00440-022-01110-7

    work page doi:10.1007/s00440-022-01110-7.url:http://dx.doi.org/10.1007/s00440-022-01110-7 2022

  3. [3]

    Ivan Corwin and Alisa Knizel.Stationary measure for the open KPZ equation. 2023. arXiv:2103. 12253 [math.PR].url:https://arxiv.org/abs/2103.12253

    work page arXiv 2023

  4. [4]

    Singular SPDEs in domains with boundaries

    Máté Gerencsér and Martin Hairer. “Singular SPDEs in domains with boundaries”. In:Probability Theory and Related Fields173.3–4 (Mar. 2018), pp. 697–758.issn: 1432-2064.doi:10 . 1007 / s00440-018-0841-1.url:http://dx.doi.org/10.1007/s00440-018-0841-1

    work page doi:10.1007/s00440-018-0841-1 2018

  5. [5]

    Derivation of the stochastic Burgers equation with Dirichlet boundary conditions from the WASEP

    Patrícia Gonçalves, Nicolas Perkowski, and Marielle Simon. “Derivation of the stochastic Burgers equation with Dirichlet boundary conditions from the WASEP”. In:Annales Henri Lebesgue3 (2020), pp. 87–167.doi:10.5802/ahl.28

    work page doi:10.5802/ahl.28 2020

  6. [6]

    Yu Gu and Jeremy Quastel.Integration by parts and invariant measure for KPZ. 2024. arXiv: 2409.08465 [math.PR].url:https://arxiv.org/abs/2409.08465

    work page arXiv 2024

  7. [7]

    Solving the KPZ equation

    Martin Hairer. “Solving the KPZ equation”. In:Annals of Mathematics178.2 (2013), pp. 559–664. doi:10.4007/annals.2013.178.2.4.url:https://annals.math.princeton.edu/2013/178- 2/p04

    work page doi:10.4007/annals.2013.178.2.4.url:https://annals.math.princeton.edu/2013/178- 2013

  8. [8]

    The snapping out Brownian motion

    Antoine Lejay. “The snapping out Brownian motion”. In:The Annals of Applied Probability26.3 (June 2016).issn: 1050-5164.doi:10.1214/15-aap1131.url:http://dx.doi.org/10.1214/15- AAP1131

    work page doi:10.1214/15-aap1131.url:http://dx.doi.org/10.1214/15- 2016

  9. [9]

    Carl Mueller and David Nualart.Regularity of the density for the stochastic heat equation. 2007. arXiv:0709.2663 [math.PR].url:https://arxiv.org/abs/0709.2663

    work page arXiv 2007

  10. [10]

    David Nualart.The Malliavin Calculus and Related Topics. 2nd. Probability and Its Applications. Berlin: Springer-Verlag, 2006.isbn: 978-3-540-28328-7

    2006

  11. [11]

    Ramiro Scorolli.Feynman-Kac formula for the heat equation driven by time-homogeneous white noise potential. 2021. arXiv:2108.12406 [math.PR].url:https://arxiv.org/abs/2108.12406

    work page arXiv 2021

  12. [12]

    Large deviations for a reaction-diffusion equation with non-Gaussian perturbations

    R. B. Sowers. “Large deviations for a reaction-diffusion equation with non-Gaussian perturbations”. In:The Annals of Probability20.2 (1992), pp. 504–537.doi:10.1214/aop/1176989797. Department of Mathematics, University of Toronto Email address:james.bonalandry@mail.utoronto.ca

    work page doi:10.1214/aop/1176989797 1992