pith. sign in

arxiv: 2607.02477 · v1 · pith:BRMYE4ZUnew · submitted 2026-07-02 · 🧮 math.PR · cs.DM· math-ph· math.MP

Mixing times of spin systems on dynamical percolation

Alexandre Stauffer , Oskar Vavtar This is my paper

Pith reviewed 2026-07-03 06:30 UTC · model grok-4.3

classification 🧮 math.PR cs.DMmath-phmath.MP
keywords mixing timesdynamical percolationGlauber dynamicsspin systemssubcritical percolationMarkov chainscoupling
0
0 comments X

The pith

Spin systems on subcritical dynamical percolation reach equilibrium in time of order log N over lambda when lambda is small.

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

The paper studies the rate at which nearest-neighbour spin systems equilibrate when their allowed interactions are confined to edges that open and close randomly over time. Edges on the torus switch independently at rate lambda according to a fixed Bernoulli probability p, while spins update at rate 1 via Glauber dynamics only across currently open edges. The central result states that whenever p lies below the percolation threshold, the mixing time of this combined non-reversible chain is of order log N divided by lambda, once lambda itself is taken sufficiently small, and this holds at any temperature. A reader would care because the result separates the fast local spin relaxation from the slower global changes in the interaction graph, showing that subcritical percolation lets the system forget its initial state on a timescale governed by the edge-update rate rather than by the volume.

Core claim

For a relatively general class of nearest-neighbour systems, as long as p is less than p_c(d), for any temperature, if lambda is sufficiently small, the mixing time of the Markov chain on the d-dimensional torus of side N is of order log N over lambda. The chain is non-reversible, and the proof proceeds by constructing a coupling that identifies local configurations whenever the dynamical percolation environment remains in a good state for long enough intervals.

What carries the argument

A coupling of local spin configurations that succeeds whenever the dynamical percolation environment remains in a good state for sufficiently long intervals.

If this is right

  • The mixing time remains logarithmic in the volume even though the chain is non-reversible.
  • The result holds uniformly in temperature and for a broad class of nearest-neighbour interactions.
  • Equilibration occurs on a timescale set by the inverse edge-update rate rather than by any volume-dependent relaxation time of the static system.
  • The same order is obtained once the percolation environment supplies sufficiently long good intervals, independent of the precise temperature.

Where Pith is reading between the lines

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

  • The same coupling strategy could be tested on other slowly evolving random environments whose connectivity stays subcritical for long stretches.
  • If lambda is not taken small, the mixing time may instead be governed by the time for the percolation configuration itself to mix.
  • Explicit tail bounds on the length of good intervals would turn the existence result into a fully quantitative statement valid for moderate lambda.

Load-bearing premise

The coupling of spin configurations succeeds whenever the percolation environment stays good for long enough intervals, without the argument supplying explicit quantitative bounds on how often or how long those intervals occur for small lambda.

What would settle it

A direct computation or simulation on a sequence of tori showing that the total-variation distance to equilibrium remains bounded away from zero after time C log N over lambda, for arbitrarily small lambda and p below threshold, would falsify the claimed upper bound on the mixing time.

read the original abstract

We study the mixing times of stochastic spin systems corresponding to nearest-neighbour Glauber dynamics on dynamical percolation, defined on $d$-dimensional torus of side-length $N$. In this model, the status of each edge (open or closed) updates independently at rate $\lambda>0$, according to $\mathrm{Ber}(p)$ samples. Simultaneously, the spin of each site updates at rate $1$ according to Glauber dynamics on the environment restricted to open edges. We show that for a relatively general class of nearest-neighbour systems, as long as $p<p_c(d)$, for any temperature, if $\lambda$ is sufficiently small, the mixing time is of order $\frac{\log N}{\lambda}$. This Markov chain is non-reversible, and the proof is obtained by developing a particular coupling that couples together local configurations whenever the environment behaves well.

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

2 major / 2 minor

Summary. The paper studies mixing times of nearest-neighbor Glauber dynamics for spin systems on dynamical percolation on the d-torus of side N, where edges flip at rate λ according to Ber(p) and spins update at rate 1 on the current open edges. For p < p_c(d), any temperature, and λ sufficiently small, it claims the mixing time is of order log N / λ. The proof uses a coupling that succeeds when the percolation environment is in a 'good' state (allowing local mixing on subcritical clusters).

Significance. If the claimed bound holds with the stated uniformity, the result would give the first explicit mixing-time control for non-reversible Glauber dynamics on time-varying random graphs in the subcritical regime, showing that sufficiently slow environment evolution (small λ) yields the same logarithmic mixing as the static subcritical case, independent of temperature. The coupling construction for time-dependent environments would be a technical contribution to the analysis of Markov chains on random media.

major comments (2)
  1. [Coupling construction and proof of mixing-time bound] The central coupling argument (developed after the model definition) requires that the dynamical percolation environment remains in a 'good' state for intervals of length ≫ log N / λ with probability 1 − o(λ / log N) uniformly in N. While subcritical tail bounds on cluster sizes are invoked, the manuscript supplies no explicit quantitative lower bound on the measure of such long good intervals after time-rescaling by λ, nor the resulting failure probability as λ → 0. This control is load-bearing for the O(log N / λ) upper bound.
  2. [Definition of good states and local mixing estimates] The definition of the 'good' environment state (used to trigger local coupling) must be accompanied by an explicit N-independent bound on the local mixing time on a good configuration; without this, the global coupling time cannot be closed at the claimed scale.
minor comments (2)
  1. [Introduction] The abstract states the result for 'a relatively general class of nearest-neighbour systems' but the precise assumptions on the single-site update rates or interaction potentials are not listed in the introduction; a short enumerated list would improve readability.
  2. [Model definition] Notation for the dynamical percolation process (edge update rate λ, occupation probability p) is introduced without an explicit comparison to the static percolation threshold p_c(d) in the model section; adding the standard reference for p_c(d) would help.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their careful reading and constructive comments, which help clarify the quantitative aspects of the coupling argument. We address the two major comments point by point below and will revise the manuscript to include the requested explicit bounds and clarifications.

read point-by-point responses

  1. Referee: [Coupling construction and proof of mixing-time bound] The central coupling argument (developed after the model definition) requires that the dynamical percolation environment remains in a 'good' state for intervals of length ≫ log N / λ with probability 1 − o(λ / log N) uniformly in N. While subcritical tail bounds on cluster sizes are invoked, the manuscript supplies no explicit quantitative lower bound on the measure of such long good intervals after time-rescaling by λ, nor the resulting failure probability as λ → 0. This control is load-bearing for the O(log N / λ) upper bound.

    Authors: We agree that the quantitative control on good intervals should be stated more explicitly. The proof uses that, after time rescaling by λ, the environment is a slow Markov chain whose stationary measure is subcritical percolation; the probability of a bad interval of length C log N / λ is at most exp(−c log N) for large C by the exponential tail on cluster sizes (P(|C_v| > k) ≤ e^{−θ k} for θ > 0 depending only on d and p < p_c). A union bound over O(1) such intervals during the mixing window then yields total failure probability o(λ / log N) uniformly in N as λ → 0. These estimates were only outlined; the revision will add a dedicated lemma giving the explicit lower bound on the measure of good intervals and the resulting failure probability. revision: yes

  2. Referee: [Definition of good states and local mixing estimates] The definition of the 'good' environment state (used to trigger local coupling) must be accompanied by an explicit N-independent bound on the local mixing time on a good configuration; without this, the global coupling time cannot be closed at the claimed scale.

    Authors: The good state is defined so that every cluster has size at most a fixed M = M(d,p) (chosen via the subcritical tail so that the probability of any larger cluster is negligible). On any such finite graph of bounded size and degree, the Glauber dynamics for the given spin system mixes in time at most T_0 = T_0(d,p,β) < ∞ independent of N; this follows from standard comparison or canonical-path arguments that depend only on the maximum degree and the inverse temperature. The revision will augment the definition of good states with this explicit N-independent bound T_0 and include a short paragraph deriving it from the cluster-size control. revision: yes

Circularity Check

0 steps flagged

No circularity: standard coupling argument yields explicit mixing-time bound

full rationale

The paper derives the O(log N / λ) mixing-time bound for small λ via a coupling construction that succeeds on intervals when the dynamical percolation environment remains subcritical. This is a direct probabilistic argument on the product chain; no parameter is fitted to data and then relabeled as a prediction, no self-citation supplies a load-bearing uniqueness theorem, and no ansatz is smuggled in. The derivation therefore stands as an independent proof rather than a tautological restatement of its inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

Only the abstract is available; the ledger is therefore populated from the model description alone. No explicit free parameters or invented entities are named.

axioms (2)
  • standard math The underlying single-site Glauber dynamics is a valid Markov chain on the spin configurations for any fixed environment.
    Invoked implicitly when the spin updates are defined at rate 1.
  • standard math The edge processes are independent continuous-time Markov chains with stationary distribution Ber(p).
    Stated directly in the model definition.

pith-pipeline@v0.9.1-grok · 5676 in / 1355 out tokens · 27006 ms · 2026-07-03T06:30:25.191520+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

30 extracted references · 5 canonical work pages · 1 internal anchor

  1. [1]

    Biased ran- dom walk on dynamical percolation.The Annals of Probability, 52(6), November 2024

    Sebastian Andres, Nina Gantert, Dominik Schmid, and Perla Sousi. Biased ran- dom walk on dynamical percolation.The Annals of Probability, 52(6), November 2024

    2024

  2. [2]

    On the stationary measures of two variants of the voter model

    Jhon Astoquillca. On the stationary measures of two variants of the voter model. Journal of Theoretical Probability, 39(2), March 2026

    2026

  3. [3]

    Sampling from Potts on random graphs of un- bounded degree via random-cluster dynamics.The Annals of Applied Probability, 33(6B), December 2023

    Antonio Blanca and Reza Gheissari. Sampling from Potts on random graphs of un- bounded degree via random-cluster dynamics.The Annals of Applied Probability, 33(6B), December 2023

    2023

  4. [4]

    Metastability of Glauber dynamics with inhomogeneous coupling disorder

    Anton Bovier, Frank den Hollander, Saeda Marello, Elena Pulvirenti, and Martin Slowik. Metastability of Glauber dynamics with inhomogeneous coupling disorder. Latin American Journal of Probability and Mathematical Statistics, 21(2):1249, 2024

    2024

  5. [5]

    Relaxation of disordered magnets in the Griffiths’ regime.Communications in Mathematical Physics, 188(1):135–173, 1997

    Filippo Cesi, Christian Maes, and Fabio Martinelli. Relaxation of disordered magnets in the Griffiths’ regime.Communications in Mathematical Physics, 188(1):135–173, 1997

    1997

  6. [6]

    The contact process can sur- vive on a slightly subcritical dynamical percolation cluster.Preprint, 2026

    Aurelia Deshayes and R´ egine Marchand. The contact process can sur- vive on a slightly subcritical dynamical percolation cluster.Preprint, 2026. arxiv:2602.19794

    work page arXiv 2026

  7. [7]

    Sander Dommers, Frank den Hollander, Oliver Jovanovski, and Francesca R. Nardi. Metastability for Glauber dynamics on random graphs.The Annals of Applied Probability, 27(4), August 2017

    2017

  8. [8]

    Logarithmic Mixing of Random Walks on Dynamical Random Cluster Models

    Andreas Galanis, Leslie Ann Goldberg, and Xandru Mifsud. Logarithmic mix- ing of random walks on dynamical random cluster models.Preprint, 2026. arxiv:2605.06511

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  9. [9]

    The random geometry of equilibrium phases.Preprint, 1999

    Hans-Otto Georgii, Olle H¨ aggstr¨ om, and Christian Maes. The random geometry of equilibrium phases.Preprint, 1999. arxiv:9905031

    1999

  10. [10]

    Random walk on dynamical percolation in Euclidean lattices: separating critical and supercritical regimes.Preprint, 2024

    Chenlin Gu, Jianping Jiang, Yuval Peres, Zhan Shi, Hao Wu, and Fan Yang. Random walk on dynamical percolation in Euclidean lattices: separating critical and supercritical regimes.Preprint, 2024. arxiv:2407.15162

    work page arXiv 2024

  11. [11]

    Speed of random walk on dynamical percolation in nonamenable transitive graphs

    Chenlin Gu, Jianping Jiang, Yuval Peres, Zhan Shi, Hao Wu, and Fan Yang. Speed of random walk on dynamical percolation in nonamenable transitive graphs. Preprint, 2024. arxiv:2407.15079

    work page arXiv 2024

  12. [12]

    Olle H¨ aggstr¨ om, Yuval Peres, and Jeffrey E. Steif. Dynamical percolation.Annales de l’Institut Henri Poincare (B) Probability and Statistics, 33(4):497–528, 1997. 34

    1997

  13. [13]

    Hayes and Alistair Sinclair

    Thomas P. Hayes and Alistair Sinclair. A general lower bound for mixing of single- site dynamics on graphs.The Annals of Applied Probability, 17(3), Jun 2007

    2007

  14. [14]

    A comparison principle for random walk on dynamical percolation.The Annals of Probability, 48(6), November 2020

    Jonathan Hermon and Perla Sousi. A comparison principle for random walk on dynamical percolation.The Annals of Probability, 48(6), November 2020

    2020

  15. [15]

    Re- sults on the contact process with dynamic edges or under renewals.Electronic Journal of Probability, 27, January 2022

    Marcelo Hil´ ario, Daniel Ungaretti, Daniel Valesin, and Maria Eul´ alia Vares. Re- sults on the contact process with dynamic edges or under renewals.Electronic Journal of Probability, 27, January 2022

    2022

  16. [16]

    Mixing time of random walk on dynam- ical random cluster.Probability Theory and Related Fields, 189(3-4):981–1043, February 2024

    Andrea Lelli and Alexandre Stauffer. Mixing time of random walk on dynam- ical random cluster.Probability Theory and Related Fields, 189(3-4):981–1043, February 2024

    2024

  17. [17]

    American Mathematical Society, October 2017

    David Levin and Yuval Peres.Markov Chains and Mixing Times. American Mathematical Society, October 2017

    2017

  18. [18]

    Liggett, Roberto H

    Thomas M. Liggett, Roberto H. Schonmann, and Alan M. Stacey. Domination by product measures.The Annals of Probability, 25(1), Jan 1997

    1997

  19. [19]

    The contact process with dynamic edges on Z.Electronic Journal of Probability, 25, January 2020

    Amitai Linker and Daniel Remenik. The contact process with dynamic edges on Z.Electronic Journal of Probability, 25, January 2020

    2020

  20. [20]

    Cutoff for the Ising model on the lattice.Inventiones mathematicae, 191(3):719–755, May 2012

    Eyal Lubetzky and Allan Sly. Cutoff for the Ising model on the lattice.Inventiones mathematicae, 191(3):719–755, May 2012

    2012

  21. [21]

    Information percolation and cutoff for the stochastic Ising model.Journal of the American Mathematical Society, 29(3):729–774, Sep 2015

    Eyal Lubetzky and Allan Sly. Information percolation and cutoff for the stochastic Ising model.Journal of the American Mathematical Society, 29(3):729–774, Sep 2015

    2015

  22. [22]

    Universality of cutoff for the Ising model.The Annals of Probability, 45(6A), Nov 2017

    Eyal Lubetzky and Allan Sly. Universality of cutoff for the Ising model.The Annals of Probability, 45(6A), Nov 2017

    2017

  23. [23]

    Cover times for random walk on dynamical percolation

    Maarten Markering. Cover times for random walk on dynamical percolation. ALEA Lat. Am. J. Probab. Math. Stat., 21(2):907–921, 2024

    2024

  24. [24]

    Biased random walk on the critical curve of dynamical percolation.Preprint, 2025

    Assylbek Olzhabayev and Dominik Schmid. Biased random walk on the critical curve of dynamical percolation.Preprint, 2025. arxiv:2502.08568

    work page arXiv 2025

  25. [25]

    Yuval Peres, Perla Sousi, and Jeffrey E. Steif. Quenched exit times for random walk on dynamical percolation.Markov Process. Related Fields, 24(5):715–731, 2018

    2018

  26. [26]

    Yuval Peres, Perla Sousi, and Jeffrey E. Steif. Mixing time for random walk on supercritical dynamical percolation.Probability Theory and Related Fields, 176(3–4):809–849, Jul 2019. 35

    2019

  27. [27]

    Yuval Peres, Alexandre Stauffer, and Jeffrey E. Steif. Random walks on dynam- ical percolation: Mixing times, mean squared displacement and hitting times. Probability Theory and Related Fields, 162(3–4):487–530, Sep 2014

    2014

  28. [28]

    Contact process in an evolving random environ- ment.Electronic Journal of Probability, 28, January 2023

    Marco Seiler and Anja Sturm. Contact process in an evolving random environ- ment.Electronic Journal of Probability, 28, January 2023

    2023

  29. [29]

    Cutoff for random walk on dynamical Erd˝ os–R´ enyi graph.Annales de l’Institut Henri Poincar´ e, Probabilit´ es et Statistiques, 56(4), Nov 2020

    Perla Sousi and Sam Thomas. Cutoff for random walk on dynamical Erd˝ os–R´ enyi graph.Annales de l’Institut Henri Poincar´ e, Probabilit´ es et Statistiques, 56(4), Nov 2020

    2020

  30. [30]

    Jeffrey E. Steif. A survey of dynamical percolation. InFractal Geometry and Stochastics IV, page 145–174. Birkh¨ auser Basel, 2009. A Validity of coupling Q f Here we verify thatQ f,g in fact defines a coupling of P f and Pg, in the sense that for eacht≥0, the marginals of the law of (σ t, ηt) and (σ ′ t, η′ t) underQ f,g correspond to P t f and Pt g resp...

    2009