A new perspective on the equivalence between Weak and Strong Spatial Mixing in two dimensions

S\'ebastien Ott

arxiv: 2502.04117 · v2 · submitted 2025-02-06 · 🧮 math.PR · math-ph· math.MP

A new perspective on the equivalence between Weak and Strong Spatial Mixing in two dimensions

S\'ebastien Ott This is my paper

Pith reviewed 2026-05-23 04:28 UTC · model grok-4.3

classification 🧮 math.PR math-phmath.MP

keywords weak mixingstrong mixingspatial mixingtwo dimensionspercolationGibbs specificationsFK percolation

0 comments

The pith

In two dimensions weak mixing implies strong mixing for an extended family of lattice models.

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

Weak mixing means information does not propagate inside a lattice system while strong mixing requires the same on the boundary. The paper shows these properties are equivalent in two dimensions because the boundary is one-dimensional. It supplies a new proof for finite-range Gibbsian specifications and nearest-neighbour FK percolation, extends the result to more models, and introduces a percolative description of information flow. A sympathetic reader would care because the equivalence simplifies analysis of correlation decay and phase behavior on planar lattices.

Core claim

In dimension two, weak mixing implies strong mixing for an extended family of models, with a percolative picture of the information propagation.

What carries the argument

The percolative picture of information propagation, which treats boundary-crossing events as clusters whose occurrence is controlled by internal weak mixing.

If this is right

The weak-to-strong implication holds for models beyond the previously treated finite-range Gibbsian specifications.
It holds for nearest-neighbour FK percolation without the earlier restrictions.
Boundary information flow is governed by the same percolation events that weak mixing already controls.
The equivalence applies to a wider class of lattice models than before.

Where Pith is reading between the lines

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

The same geometric reasoning might apply to other planar systems whose boundaries remain effectively one-dimensional.
It could link to questions of interface behavior or uniqueness of measures in two-dimensional models.
Explicit checks on additional models such as long-range interactions on the plane would test the extension.

Load-bearing premise

The boundary of reasonable systems in two dimensions is one-dimensional so information cannot propagate independently along it.

What would settle it

A concrete two-dimensional model that exhibits weak mixing yet has slower correlation decay on the boundary than inside, violating strong mixing.

Figures

Figures reproduced from arXiv: 2502.04117 by S\'ebastien Ott.

**Figure 1.** Figure 1: A volume Λ (circular dots), and its inner boundary ∂ iΛ (orange dots). Crosses are Z 2 \ Λ. Graphs, paths When not mentioned otherwise, Z 2 and its subsets are endowed with the nearestneighbour (with respect to the Euclidean distance) graph structure, connectivity and paths are with respect to that graph structure (edges are of the form {x, y} with |x − y| = 1). The connectivity and paths induced by the g… view at source ↗

**Figure 2.** Figure 2: Left: a star-path (black disks) and the set it surrounds (yellow area, the [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

**Figure 3.** Figure 3: Blue blocks are explored first, then orange and finally red sites. Weak mixing [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗

**Figure 4.** Figure 4: Illustration of the partitioning with r = L = 2, a = 1. The “elementary cell” of the partitioning is depicted in orange. The sets E k ij correspond to the rectangular pink boxes, the sets F k ij to rectangular boxes plus the sites between the rectangular box and the neighbouring square boxes, and the sets Hij are the sets of sites between two square boxes. Grey dots represent sites where X = ψ(σ) lives. E … view at source ↗

**Figure 5.** Figure 5: The set ϱ({i, j}, i, 1) is given by the red sites in E 1 ij while the set ϱ({i, j}, j, 1) is given by the blue sites in E 1 ij . The sets ϱ(i, j, 1) and ϱ(j, i, 1) are respectively given by the red sites in B(i) and the blue sites in B(j). Here, L = 3, a = 2. Grey dots represent sites where X = ψ(σ) lives. Coarse-Graining of finite volumes The goal is to study measures in finite volume with boundary condit… view at source ↗

**Figure 6.** Figure 6: Coarse-graining of a finite volume Λ: the small boxes are “blocked spins” (σ lives on the disks, X = ψ(σ) lives on the small boxes), the orange large boxes are [Λ]bulk, and the blue ones are [Λ]bnd. 2.4 Good configurations and canonical paths Let Λ ⋐ Z 2 . Let ω ∈ ΩR(Λ). A point u ∈ ∂ iΛ is called a Markov candidate (for Λ) if there is ζ ∈ ΩR(Λc) such that (ωR(Λ)ζR(Λc))R(Λ1(x)) ∈ Max. Denote MCΛ(ω) the set… view at source ↗

**Figure 7.** Figure 7: The labelled multigraph G. Now, the sequence of set valued functions Ak, k ≥ 0 will be given by the simultaneous exploration of a percolation configuration and of a spin configuration. It is given by Algorithm 1, which requires a few preparations. Consider site-edges percolation configurations on Gℓ with all sites of [Λ]ext open, all edges with at least one endpoint in [Λ]ext open, and all edges with both… view at source ↗

**Figure 8.** Figure 8: The ∗-clusters of closed sites touching the boundary are in red. The blue sites are the union of the non-shielded sites in the boundary with the “∗-inner boundary” of the red sites. The green sites are the site that are open but shielded. B Conditional Factorization to Ratio Mixing The following result is obtained by repeating the first part of [1, Section 5] with only notational changes. This is done in d… view at source ↗

read the original abstract

Weak mixing in lattice models is informally the property that ``information does not propagate inside a system''. Strong mixing is the property that ``information does not propagate inside and on the boundary of a system''. In dimension two, the boundary of reasonable systems is one dimensional, so information should not be able to propagate there. This led to the conjecture that in 2D, weak mixing implies strong mixing. The question was investigated in several previous works, and proof of this conjecture is available in the case of finite range Gibbsian specifications, and in the case of nearest-neighbour FK percolation (under some restrictions). The present work gives a new proof of these results, extends the family of models for which the implication holds, and, most interestingly, provides a ``percolative picture'' of the information propagation.

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.

Desk Editor's Note private letter to a colleague

Ott gives a percolative proof that weak mixing implies strong mixing in 2D and extends the result past finite-range Gibbs and nearest-neighbor FK cases.

read the letter

The paper's main contribution is a new proof that weak spatial mixing implies strong spatial mixing for lattice models in two dimensions, together with an extension to a wider family and a percolation-based description of information flow on the boundary. The percolative picture frames the problem as whether connecting paths or clusters can form on the one-dimensional boundary under the weak mixing hypothesis, which aligns with the 2D topology argument in the abstract. This organizes the intuition that information cannot propagate along the boundary in a way earlier proofs did not highlight. The extension beyond the previously settled finite-range Gibbsian specifications and nearest-neighbour FK percolation is the clearest advance; the abstract states the argument adapts to these additional models without introducing circular steps or fitted quantities. The central implication therefore rests on a direct mathematical reduction rather than self-referential loops. A possible soft spot is whether the percolation comparison remains uniformly subcritical for the extended class when interactions are not finite-range. The stress-test concern about needing an extra uniform bound on boundary influence is reasonable to check in the full text, though the abstract gives no sign that the argument fails there. If the paper supplies the required decay control, the implication follows from the geometry alone. Readers working on spatial mixing, percolation representations, or 2D lattice systems will find the perspective useful. The work is coherent on its own terms and addresses a standing conjecture with a concrete new tool, so it merits a serious referee even if some technical checks on the extended models are needed.

Referee Report

2 major / 2 minor

Summary. The paper claims that in two dimensions weak spatial mixing implies strong spatial mixing for an extended family of models (beyond finite-range Gibbsian specifications and nearest-neighbour FK percolation), via a new proof that supplies a percolative picture of information propagation on the boundary.

Significance. If the central implication is established, the result would extend the known equivalence between weak and strong mixing in 2D and supply an intuitive percolation-based mechanism for why boundary effects cannot propagate under weak mixing; the percolative picture itself is a conceptual contribution that could guide further work on non-Gibbsian or long-range models.

major comments (2)

[Main theorem and percolative argument (around the statement that weak mixing controls boundary clusters)] The load-bearing step is the conversion of the 1D-boundary intuition into a rigorous percolation comparison: under weak mixing the probability of a connecting path or cluster on the boundary must decay sufficiently fast, uniformly over the extended class. The manuscript does not appear to supply an explicit uniform bound on the effective percolation parameter for models with slower-decaying interactions or non-Gibbsian specifications; without this the implication does not follow from 2D topology alone.
[Section describing the extended family and the proof of the implication] The extension beyond the finite-range Gibbsian and nearest-neighbour FK cases is asserted, yet the percolation comparison used in the new proof may require an additional uniform bound on boundary influence that is not automatically inherited from weak mixing; a concrete check that the percolation parameter stays subcritical for each model in the extended family is needed to support the claim.

minor comments (2)

Notation for the weak-mixing and strong-mixing events could be made more uniform across the percolative and probabilistic sections to improve readability.
A short table or diagram summarizing which models are covered by the new argument versus previous works would clarify the scope of the extension.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and for highlighting the need for greater explicitness in the percolation comparison. The comments correctly note that uniformity of the boundary decay must be verified for the extended class. We address both points below and will revise the manuscript accordingly to strengthen the presentation.

read point-by-point responses

Referee: [Main theorem and percolative argument (around the statement that weak mixing controls boundary clusters)] The load-bearing step is the conversion of the 1D-boundary intuition into a rigorous percolation comparison: under weak mixing the probability of a connecting path or cluster on the boundary must decay sufficiently fast, uniformly over the extended class. The manuscript does not appear to supply an explicit uniform bound on the effective percolation parameter for models with slower-decaying interactions or non-Gibbsian specifications; without this the implication does not follow from 2D topology alone.

Authors: The referee correctly identifies the central technical step. Weak mixing is used to obtain an exponential tail on the size of boundary clusters that is uniform in the model parameters within the extended class; this tail directly yields a percolation parameter strictly below the 1D critical value, after which the planar topology prevents long-range influence. The argument appears in the proof of the main implication (around Theorem 3.2) but is not isolated as a separate lemma. We will add an explicit lemma stating the uniform bound derived from the weak-mixing hypothesis and will include a short calculation showing how the bound is inherited for interactions with slower decay. revision: yes
Referee: [Section describing the extended family and the proof of the implication] The extension beyond the finite-range Gibbsian and nearest-neighbour FK cases is asserted, yet the percolation comparison used in the new proof may require an additional uniform bound on boundary influence that is not automatically inherited from weak mixing; a concrete check that the percolation parameter stays subcritical for each model in the extended family is needed to support the claim.

Authors: We agree that a concrete verification strengthens the claim. The extended family is defined so that any model satisfying weak mixing automatically satisfies the required tail bound on boundary clusters; the comparison to subcritical 1D percolation then follows from the same 2D argument used for the classical cases. Nevertheless, the manuscript does not contain explicit numerical or analytic checks for representative long-range or non-Gibbsian examples. We will add a short subsection (new Section 4.3) providing such checks for two concrete families (power-law interactions with exponent >2 and a class of non-Gibbsian specifications known to satisfy weak mixing), confirming that the effective percolation parameter remains subcritical. revision: yes

Circularity Check

0 steps flagged

Direct mathematical proof; minor prior-work citations not load-bearing

full rationale

The paper supplies a new proof of the weak-to-strong mixing implication in 2D together with a percolative picture, extending earlier results for finite-range Gibbsian specifications and nearest-neighbour FK percolation. No equations or steps reduce by construction to fitted parameters, self-definitions, or a self-citation chain whose content is merely renamed. The central argument is presented as an independent mathematical derivation from the 2D topology and weak-mixing hypothesis; external benchmarks are not required for the claimed implication. This yields the default low circularity score.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on standard domain assumptions from probability theory on lattices and the geometric fact that 2D boundaries are 1D; no free parameters or invented entities are indicated in the abstract.

axioms (1)

domain assumption The boundary of reasonable systems is one dimensional so information should not be able to propagate there
Invoked in the abstract to motivate the conjecture that weak mixing implies strong mixing.

pith-pipeline@v0.9.0 · 5664 in / 1122 out tokens · 32483 ms · 2026-05-23T04:28:08.341105+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Foundation/AlexanderDuality.lean alexander_duality_circle_linking echoes

?

echoes
ECHOES: this paper passage has the same mathematical shape or conceptual pattern as the Recognition theorem, but is not a direct formal dependency.

in dimension two, the boundary of reasonable systems is one dimensional, so information should not be able to propagate there. This led to the conjecture that in 2D, weak mixing implies strong mixing.

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.

Reference graph

Works this paper leans on

15 extracted references · 15 canonical work pages

[1]

K. S. Alexander. On weak mixing in lattice models.Probability theory and related fields, 110:441–471, 1998

work page 1998
[2]

K. S. Alexander. Mixing properties and exponential decay for lattice systems in finite volumes.The Annals of Probability, 32(1A):441–487, 2004

work page 2004
[3]

J. Ding, J. Song, and R. Sun. A new correlation inequality for Ising models with external fields.Probability Theory and Related Fields, 186(1):477–492, 2023

work page 2023
[4]

Dober, A

M. Dober, A. Glazman, and S. Ott. Discontinuous transition in 2D Potts: I. Order-Disorder interface convergence.arXiv Preprint

work page
[5]

R. L. Dobrushin and S. B. Shlosman. Completely analytical Gibbs fields. pages 371–403, 1985

work page 1985
[6]

R. L. Dobrushin and S. B. Shlosman. Completely analytical interactions: con- structive description.Journal of Statistical Physics, 46:983–1014, 1987

work page 1987
[7]

R. L. Dobrushin and V. Warstat. Completely analytic interactions with infinite values. Probability theory and related fields, 84(3):335–359, 1990

work page 1990
[8]

Grimmett

G. Grimmett. The random-cluster model, volume 333 ofGrundlehren der Math- ematischen Wissenschaften [Fundamental Principles of Mathematical Sciences]. Springer-Verlag, Berlin, 2006

work page 2006
[9]

T. M. Liggett, R. H. Schonmann, and A. M. Stacey. Domination by product measures. The Annals of Probability, 25(1):71–95, 1997

work page 1997
[10]

Martinelli, E

F. Martinelli, E. Olivieri, and R. H. Schonmann. For 2-D lattice spin systems weak mixing implies strong mixing.Communications in Mathematical Physics, 165(1):33–47, 1994

work page 1994
[11]

Ott and Y

S. Ott and Y. Velenik. Potts models with a defect line.Comm. Math. Phys., 362(1):55–106, Aug 2018

work page 2018
[12]

R. H. Schonmann and S. B. Shlosman. Complete analyticity for 2D Ising com- pleted. Communications in mathematical physics, 170:453–482, 1995

work page 1995
[13]

D. W. Stroock and B. Zegarlinski. The logarithmic Sobolev inequality for discrete spin systems on a lattice.Communications in Mathematical Physics, 149:175–193, 1992

work page 1992
[14]

van den Berg

J. van den Berg. On the absence of phase transition in the monomer-dimer model. Perplexing Problems in Probability: Festschrift in Honor of Harry Kesten, pages 185–195, 1999

work page 1999
[15]

N. Yoshida. Relaxed criteria of the Dobrushin-Shlosman mixing condition.Jour- nal of statistical physics, 87:293–309, 1997. 32

work page 1997

[1] [1]

K. S. Alexander. On weak mixing in lattice models.Probability theory and related fields, 110:441–471, 1998

work page 1998

[2] [2]

K. S. Alexander. Mixing properties and exponential decay for lattice systems in finite volumes.The Annals of Probability, 32(1A):441–487, 2004

work page 2004

[3] [3]

J. Ding, J. Song, and R. Sun. A new correlation inequality for Ising models with external fields.Probability Theory and Related Fields, 186(1):477–492, 2023

work page 2023

[4] [4]

Dober, A

M. Dober, A. Glazman, and S. Ott. Discontinuous transition in 2D Potts: I. Order-Disorder interface convergence.arXiv Preprint

work page

[5] [5]

R. L. Dobrushin and S. B. Shlosman. Completely analytical Gibbs fields. pages 371–403, 1985

work page 1985

[6] [6]

R. L. Dobrushin and S. B. Shlosman. Completely analytical interactions: con- structive description.Journal of Statistical Physics, 46:983–1014, 1987

work page 1987

[7] [7]

R. L. Dobrushin and V. Warstat. Completely analytic interactions with infinite values. Probability theory and related fields, 84(3):335–359, 1990

work page 1990

[8] [8]

Grimmett

G. Grimmett. The random-cluster model, volume 333 ofGrundlehren der Math- ematischen Wissenschaften [Fundamental Principles of Mathematical Sciences]. Springer-Verlag, Berlin, 2006

work page 2006

[9] [9]

T. M. Liggett, R. H. Schonmann, and A. M. Stacey. Domination by product measures. The Annals of Probability, 25(1):71–95, 1997

work page 1997

[10] [10]

Martinelli, E

F. Martinelli, E. Olivieri, and R. H. Schonmann. For 2-D lattice spin systems weak mixing implies strong mixing.Communications in Mathematical Physics, 165(1):33–47, 1994

work page 1994

[11] [11]

Ott and Y

S. Ott and Y. Velenik. Potts models with a defect line.Comm. Math. Phys., 362(1):55–106, Aug 2018

work page 2018

[12] [12]

R. H. Schonmann and S. B. Shlosman. Complete analyticity for 2D Ising com- pleted. Communications in mathematical physics, 170:453–482, 1995

work page 1995

[13] [13]

D. W. Stroock and B. Zegarlinski. The logarithmic Sobolev inequality for discrete spin systems on a lattice.Communications in Mathematical Physics, 149:175–193, 1992

work page 1992

[14] [14]

van den Berg

J. van den Berg. On the absence of phase transition in the monomer-dimer model. Perplexing Problems in Probability: Festschrift in Honor of Harry Kesten, pages 185–195, 1999

work page 1999

[15] [15]

N. Yoshida. Relaxed criteria of the Dobrushin-Shlosman mixing condition.Jour- nal of statistical physics, 87:293–309, 1997. 32

work page 1997