Finitary codings for gradient models and a new graphical representation for the six-vertex model

Gourab Ray; Yinon Spinka

REVIEW 2 major objections 3 minor 41 references

Finitary codings for gradient models and a new graphical representation for the six-vertex model

T0 review · 2 major / 3 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Global spin information, not local disorder, is the only barrier to coding low-temperature spin models finitarily from i.i.d. randomness.

desk verdict Strong new results on finitary codings of gradients, but Theorem 1.4's six-vertex threshold has a real factor-3 inconsistency that the authors should fix before publication. read the letter →

arxiv 1908.09056 v2 pith:NYVTXM7Y submitted 2019-08-24 math.PR

Gourab Ray , Yinon Spinka This is my paper · ORCID

classification math.PR MSC 60K3582B2082B43

keywords finitaryfactorofi.i.d.IsingmodelPottssix-vertexrandom-clustersuperimposedgradientcodingradius

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper shows that the only obstruction to coding low-temperature spin models finitarily from i.i.d. randomness is their global spin information. For the Ising model below critical temperature, the gradient—the edges whose endpoints have opposite spins—is a finitary factor of i.i.d. with exponentially decaying coding radius at every inverse temperature, and similarly for Potts and beach model gradients once the underlying random-cluster measures have a unique Gibbs state. For the six-vertex F-model at large parameter $c$, the height function's gradient itself remains an obstruction, but the absolute value of its diagonal gradient and of its Laplacian are finitary factors. The proof supplies a general cluster-tree mechanism and a new graphical representation of the six-vertex model.

What carries the argument

The load-bearing object is a deterministic cluster-tree factor map: for any percolation configuration with a unique infinite cluster, each finite cluster is assigned a parent (the largest-diameter cluster at a large enough scale), so the collection of clusters forms a tree rooted at the infinite cluster, and the tree-path between any two vertices' clusters can be recovered by a finite exploration. This turns independently colored clusters into a gradient by reading spin differences along tree paths, and it is what upgrades 'gradient of independently colored clusters' to a finitary factor. The paper's second mechanism is the superimposed random-cluster model—two random-cluster configurations, one on the even sublattice and one on its dual, with no closed primal-dual cross—which is monotone under an order reversed on one sublattice, has a unique Gibbs state for large $\alpha$, and couples to the six-vertex spin representation with $c=2+\alpha$.

What would settle it

Find a percolation process on a transitive graph with a unique infinite cluster and exponentially decaying cluster-size, distance-to-infinite-cluster, and connectivity tails for which the gradient of independently colored clusters is not a finitary factor, or show that the monotone-coding theorem fails under a reversed partial order. Either would break the chain behind Theorem 1.1, 1.3, 1.4, or 1.5.

Watch

Extended reading notes

Core claim

The central discovery is that global transformations—a global spin flip in Ising, a global permutation in Potts and beach models—are exactly what prevents finitary codability, and that discarding this global information through a local gradient leaves a field that is finitarily codable from i.i.d. input. For the low-temperature Ising model, the paper proves the gradient percolation configuration is ffiid with exponential tails on the coding radius for every $\beta>\beta_c(d)$, and derives a volume-order large deviation estimate for the energy. For the six-vertex F-model with $c>2+p_c/(1-p_c)$, it proves the height function's diagonal-gradient absolute value and Laplacian absolute value are ffiid, while the gradient and Laplacian themselves are not even $(\mathbb{Z}^2)_{\mathrm{even}}$-ffiid; the mechanism is a new Edwards-Sokal-type coupling between the six-vertex model and a superimposed pair of random-cluster models on primal and dual lattices.

Load-bearing premise

The proof relies on a theorem stating that a monotone model whose extremal measures coincide is a finitary factor of i.i.d.; in the six-vertex case the paper uses a version of this theorem for a partial order reversed on one sublattice, with only a two-sentence sketch, so if that theorem or its reversed-order extension fails, the positive ffiid results do not follow.

Editorial extensions

If this is right

The Ising energy per edge satisfies an exponential large-deviation bound in every dimension $d\ge 2$ and at every positive temperature, including criticality, because the gradient is ffiid and ffiid fields obey the ergodic theorem at exponential rate.
The gradient of the $q$-state Potts model is ffiid exactly when the free and wired random-cluster measures coincide, giving a concrete condition under which low-temperature Potts behavior is finitarily codeable.
The beach-model gradient inherits ffiid from uniqueness of the beach-random-cluster Gibbs state.
The six-vertex absolute diagonal gradient and absolute Laplacian are ffiid for $c>2+p_c/(1-p_c)$, while the height gradient itself is not $(\mathbb{Z}^2)_{\mathrm{even}}$-ffiid.
The superimposed random-cluster model provides a new graphical representation of the six-vertex model with $c\ge 2$, coupling six-vertex correlations to the connectivity of two random-cluster layers.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The same cluster-tree construction should apply to any percolation process with a unique infinite cluster and exponential cluster-size, distance, and connectivity tails, so the method likely extends to other models with global symmetries, such as hard-core even/odd states, once a suitable gradient is defined.
If the superimposed model's uniqueness could be pushed below the threshold $\alpha>3p_c/(1-p_c)$, the six-vertex ffiid result would extend to smaller $c$; the paper identifies this as open, and the coupling suggests a natural route through subcritical disagreement percolation.
Because the superimposed model at $\alpha=0$ coincides with the critical random-cluster model with $q=4$, a deeper coupling between the two representations might transfer known random-cluster results to six-vertex problems.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The paper proves something genuinely new: the gradient of the low-temperature Ising model is ffiid at every beta, with exponential tails, and similarly for Potts and beach gradients under the right random-cluster uniqueness condition. The new superimposed random-cluster model for the six-vertex model is a real contribution in its own right, and the cluster-tree construction in Section 2 is an elegant general tool. The main arc of the argument is credible and detailed.

The soft spots are real but mostly minor. The most concrete issue is the threshold in Theorem 1.4. The statement says c > 2 + pc/(1-pc), but the proof in Sections 4 and 5 requires α > 3pc/(1-pc), i.e. c > 2 + 3pc/(1-pc). The abstract's c ≳ 6.4 matches the proof, so the theorem statement appears to have lost a factor 3. This is a checkable inconsistency in a central claim, though it does not undermine the qualitative result—the paper still proves ffiid for all c above about 6.4. It should be fixed before publication, either by restating the theorem at the stronger threshold or by finding an argument that actually works at the weaker one.

The other soft spot is the use of the Harel–Spinka monotone-coding theorem in the reversed partial order. That is a legitimate external input, not circular, but the adaptation is only sketched in two sentences and deserves a fuller proof. The disagreement-percolation and exploration arguments are mostly standard, though compressed.

Overall, the paper is a solid piece of work with substantial new theorems and a new graphical representation. It deserves a serious referee. I would send it to peer review, ask the authors to correct the threshold and expand the reverse-order coding argument, and then accept after those revisions.

Referee Report

2 major / 3 minor

Summary. The paper studies finitary factors of i.i.d. fields for gradient-type observables of spin systems with multiple Gibbs states. The main theorems state: (Theorem 1.1) the edge percolation of unequal spins in the low-temperature Ising model is ffiid with exponential coding radius for all β>β_c(d); (Theorem 1.3) the oriented-edge gradient of the q-state Potts model is ffiid if and only if the free and wired random-cluster measures coincide; (Theorem 1.5) the gradient of the beach model is ffiid under the analogous uniqueness condition; and (Theorem 1.4) for the six-vertex/F-model height function with flat boundary conditions and c>2+p_c/(1−p_c), the fields h, ∇h, ∇_d h, and Δh are not (Z2)even-ffiid, while |∇_d h| and |Δh| are ffiid. The proofs introduce a universal cluster-tree construction on percolation clusters (Section 2) and a new 'superimposed random-cluster model' coupled to the six-vertex model (Section 4). The six-vertex positive results are proved only under the stronger threshold c>2+3p_c/(1−p_c), as detailed in major comment 1.

Significance. The overarching message—that global spin-flip or spin-permutation information is the only obstruction to ffiid, and that suitable gradients are ffiid—is natural and, if fully established, would be a substantial contribution to the finitary-coding literature. The paper's concrete assets include a deterministic cluster-tree factor map that is universal across percolation processes with a unique infinite cluster, and a new graphical representation of the six-vertex model that is likely to be useful independently. The results are stated as theorems proved from explicit assumptions rather than inferred from fitted parameters, and the main proof strategy is coherent. However, the factor-3 threshold gap in Theorem 1.4 means that one of the central positive claims is currently established on a smaller parameter region than the theorem statement asserts. The reversed-order application of the Harel–Spinka theorem is also only sketched. With these points addressed, the paper would be a strong addition to the field.

major comments (2)

[Theorem 1.4; opening line of Section 5; Lemma 4.10; Theorem 4.7] The parameter threshold in Theorem 1.4 is not the threshold used in the proof. The theorem states c > 2 + p_c/(1−p_c), and Section 5 repeats 'fix c > 2 + p_c/(1−p_c) and α = c−2'. However, Theorem 4.7 requires α > (p_c/(1−p_c)) max{q+1,2}, which for q=2 is α > 3p_c/(1−p_c). Equivalently, Lemma 4.10 uses a dominated Bernoulli site percolation with parameter p = α/(max{2,q+1}+α), and the condition p > p_c is exactly α > 3p_c/(1−p_c) when q=2. Proposition 4.11 and Corollary 4.12 likewise fix α > 3p_c/(1−p_c). Since α = c−2, the supplied arguments establish Theorem 1.4 only for c > 2 + 3p_c/(1−p_c) (approximately 6.365), not for c > 2 + p_c/(1−p_c) (approximately 3.455). The abstract's 'c ≳ 6.4' is consistent with the proof, so the theorem statement should be corrected to the stronger threshold or the argument must be extended to the claimed range.
[Section 5, proof of Proposition 5.2] The ffiid property of the superimposed model is obtained by applying [29, Theorem 7] to a partial order that is reversed on one sublattice, with only a two-sentence remark that the proof of that theorem extends to this setting. Because the positive directions of Theorems 1.1, 1.3, 1.4, and 1.5 depend on monotone-coding inputs, this adaptation is load-bearing. I recommend that the authors state and prove the reversed-order version as a lemma, or provide a complete verification that [29, Theorem 7] applies verbatim to the partial order in (4.8). As written, the argument relies on an unproved adaptation of an external theorem.

minor comments (3)

[Section 1.5, six-vertex outline] The sentence 'the six-vertex model can be coupled with the six-vertex model' appears to contain a typo; it should refer to the superimposed random-cluster model.
[References] Reference [21] is listed as 'In preparation, 2019'; the published or updated version should be cited if available.
[Section 4.1, equation (4.1)] The definition of kΔ(η) as 'the sum kΔ(η0)+kΔ(η1) of the number of open vertex-clusters' is somewhat redundant; consider simplifying to 'the number of open clusters in η0 plus the number in η1 that meet Δ'.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: all target ffiid claims are derived from explicit cluster-tree constructions and from external monotone-coding theorems whose assumptions do not include the target results.

full rationale

The derivation chain is not circular. Theorem 1.3 reduces the Potts-gradient claim to the random-cluster model: sufficiency invokes [29, Thm 1] for ffiid of the RC measure when free and wired measures coincide, then applies the self-contained cluster-tree coding of Theorem 2.2; necessity proves that an ffiid gradient would make the wired RC measure ffiid by Bernoulli percolation on constant-spin clusters, contradicting the independent non-ffiid criterion [29, Thm 2]. Theorem 1.1 uses the same route with the known coincidence of FK-Ising measures, and Theorem 1.5 uses the monotone beach-RC theorem [29, Thm 7] together with Theorem 2.2. For the six-vertex model, the superimposed random-cluster model is introduced and coupled to the spin representation in Proposition 4.1; its uniqueness for large alpha is proved in Theorem 4.7 by disagreement percolation, not assumed. The ffiid step for the superimposed model is imported from [29, Thm 7]; although that theorem is by the same second author, its assumptions (monotone model, coinciding extremal measures) do not include the target gradient/Laplacian claim, so this is legitimate reliance on separate prior work rather than circularity. The reverse-order adaptation of [29, Thm 7] is sketched in two sentences and is therefore an external dependency or correctness risk, but it is not a reduction of the conclusion to itself. Likewise, the threshold inconsistency flagged by the skeptic (Theorem 1.4 states c > 2 + p_c/(1-p_c), whereas Theorem 4.7 and Proposition 4.11 require alpha > 3 p_c/(1-p_c), i.e. c > 2 + 3p_c/(1-p_c)) is a numerical gap in the written proof, not a circular step: the proof and the theorem statement are not equivalent by construction, they merely disagree about a constant. No fitted parameter is relabelled as a prediction, no uniqueness theorem is imported to forbid alternatives without proof, and no known result is renamed as a new object. Overall, the paper is self-contained on its coding constructions and uses its cited theorems as genuine external evidence.

Assumptions & free parameters 0 free parameters · 5 assumptions · 1 invented entities

The central results are new theorems built on recent finitary-coding results of the second author and collaborators. The main external assumptions are the monotone-model ffiid theorem and the non-ffiid criteria for random-cluster measures; both are cited from arXiv preprints rather than proved here. No free parameters are fitted. One new object, the superimposed model, is introduced as a tool.

assumptions (5)

domain assumption Monotone models with coinciding extremal measures are ffiid (Harel-Spinka theorem).
Invoked to show random-cluster, beach-random-cluster, and superimposed models are ffiid; not proved in the paper, and the superimposed case needs a reverse-order adaptation.
domain assumption If free and wired random-cluster measures differ, the random-cluster model is not ffiid (Harel-Spinka and Spinka).
Used for the negative direction of Theorem 1.3 and for the non-ffiid parts of Theorem 1.4.
domain assumption Free and wired FK-Ising (q=2) random-cluster measures coincide, with quantitative mixing and good-box estimates.
Used in the proof of Theorem 1.1 to get exponential tails of the coding radius; not proved in the paper.
standard math Standard facts about Bernoulli site percolation on Z^2, including pc and exponential decay in the subcritical regime.
Used in Theorem 4.7, Lemma 4.10, and Proposition 4.11 for the disagreement-percolation argument and circuit estimates.
domain assumption van den Berg-Steif theorem: Ising plus state is ffiid iff unique Gibbs measure; and the exponential ergodic theorem for ffiid fields.
Provides context and is used in Corollary 1.2 and Proposition 5.5.

invented entities (1)

Superimposed random-cluster model
purpose: Graphical representation of the six-vertex model with c>=2, coupling the F-model spin representation to a pair of primal and dual random-cluster percolations via an Edwards-Sokal-like coupling.
Introduced in Section 4; its properties (FKG, uniqueness for large alpha, unique infinite clusters) are proved inside the paper. No external falsifiable handle beyond the paper itself.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Finitary codings for gradient models and a new graphical representation for the six-vertex model." pith.science (2026). https://pith.science/paper/NYVTXM7Y

@misc{pith2026190809056,
  author       = {Pith},
  title        = {Pith review of: Finitary codings for gradient models and a new graphical representation for the six-vertex model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NYVTXM7Y}},
  note         = {Machine review of arXiv:1908.09056}
}

abstract

It is known that the Ising model on $\mathbb {Z}^d$ at a given temperature is a finitary factor of an i.i.d. process if and only if the temperature is at least the critical temperature. Below the critical temperature, the plus and minus states of the Ising model are distinct and differ from one another by a global flip of the spins. We show that it is only this global information which poses an obstruction for being finitary by showing that the gradient of the Ising model is a finitary factor of i.i.d. at all temperatures. As a consequence, we deduce a volume-order large deviation estimate for the energy. A similar result is shown for the Potts model. A result in the same spirit is also shown for the six-vertex model, which is itself the gradient of a height function, with parameter $c \gtrapprox 6.4$. We show that the gradient of the height function is not a finitary factor of an i.i.d. process, but that its "Laplacian" is. For this, we introduce a coupling between the six-vertex model with $c\ge 2$ and a new graphical representation of it, reminiscent of the Edwards--Sokal coupling between the Potts and random-cluster models. We believe that this graphical representation may be of independent interest and could serve as a tool in further understanding of the six-vertex model. To provide further support for the ubiquity of this type of phenomenon, we also prove an analogous result for the so-called beach model. The tools and techniques used in this paper are probabilistic in nature. The heart of the argument is to devise a suitable tree structure on the clusters of the underlying percolation process (associated to the graphical representation of the given model), which can be revealed piece-by-piece via exploration.

Figures

Figures reproduced from arXiv: 1908.09056 by the authors.

**Figure 1.** The six types of arrow configurations satisfying the ice rule at a vertex, and the corresponding height function (which is assumed in the figure to be 0 on the bottom-left face). for an oriented edge (u, v) ∈ E~ (Z d ). A Gibbs measure for the Potts model induces a probability measure on {0, . . . , q − 1} E~ (Z d) via the map σ 7→ ∇σ. We call the measure induced by a constant boundary condition Gibbs state (any co… view at source ↗

**Figure 2.** A height function and its corresponding six-vertex configuration on a diamond domain Λ with 0 boundary condition (meaning that the height is fixed to be 0 and 1 on the internal and external vertex boundaries). The black circuit is ∂ †Λ. The gradient of a height function h lives on the oriented edges of (Z 2 ) ∗ and is defined by (∇h)(u,v) = h(u) − h(v) (1.3) for an oriented edge (u, v) ∈ E~ ((Z 2 ) ∗ ). We note that… view at source ↗

**Figure 3.** The spin representation of the six-vertex model. The spin is + if the height is 0 or 1 modulo 4, and − otherwise. The height is assumed in the figure to be 0 on the bottom-left square. If instead it were assumed to be 2, then the spins would be globally flipped. Let us consider next the beach model (Theorem 1.5). Using a general result from [29] about finitary codings for monotone models, we obtain that the beach-ra… view at source ↗

Figures from the paper (5 more)

**Figure 4.** Figure 4: A superimposed configuration η with a compatible spin configuration σ. The blue and red circles correspond to − and + spins in σ, respectively. Sites which are connected by edges of η must have the same spin value in σ. Unlike in the usual random-cluster model, an edge…

**Figure 5.** Figure 5: An illustration of the cluster-tree construction. Each finite cluster has an arrow pointing to its parent. Given two vertices u and v, one may find in a finitary manner the paths (in dark gray) in the cluster-tree from the clusters of u and v (in orange) to their lowes…

**Figure 6.** Figure 6: An illustration of how the parent of a finite cluster C is found in a finitary manner. In both situations depicted, d = diam C, k = k(C), the parent of C is shown in blue and a ball around C witnessing its parent is shown in light green. The dark gray clusters were tes…

**Figure 7.** Figure 7: A superimposed configuration with wired-wired boundary conditions and a compatible spin configuration with +− boundary conditions (+ in red, − in blue) on a diamond domain. Proposition 4.1. Let c > 2 and set α = c − 2. Let i, j ∈ {+, −} and let Λ be a diamond domain an…

**Figure 8.** Figure 8: Proof of Lemma 4.9. The solid discs denote X and the black path is the circuit C. The primal and dual edges formed by the open crosses in the vertices of C are drawn in blue and red, respectively. Note that X might not be ∗-connected in Z 2 . Proof. Let η ∈ Ω SI,τ ∆ an…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

41 extracted references · 36 canonical work pages

[1]

S. Adams. Følner independence and the amenable Ising model. Ergodic Theory and Dynamical Systems, 12(4):633–657, 1992

work page 1992
[2]

Aizenman, H

M. Aizenman, H. Duminil-Copin, and V. Sidoravicius. Random currents and continuity of Ising models spontaneous magnetization. Communications in Mathematical Physics , 334(2):719–742, 2015

work page 2015
[3]

Emergent Planarity in two-dimensional Ising Models with finite-range Interactions

M. Aizenman, H. Duminil-Copin, V. Tassion, and S. Warzel. Emergent planarity in two-dimensional ising models with ﬁnite-range interactions. Inventiones, to appear, 2018. arXiv:1801.04960

work page Pith review arXiv 2018
[4]

R. J. Baxter, S. B. Kelland, and F. Y. Wu. Equivalence of the Potts model or Whitney polynomial with an ice-type model. Journal of Physics A: Mathematical and General , 9(3):397, 1976

work page 1976
[5]

J. v. d. Berg and C. Maes. Disagreement percolation in the study of Markov ﬁelds. The Annals of Probability, 22(2):749–763, 1994

work page 1994
[6]

J. v. d. Berg and J. E. Steif. On the existence and nonexistence of ﬁnitary codings for a class of random ﬁelds. Annals of probability, pages 1501–1522, 1999

work page 1999
[7]

Bodineau

T. Bodineau. Slab percolation for the Ising model. Probab. Theory Related Fields, 132(1):83–118, 2005

work page 2005
[8]

Bodineau

T. Bodineau. Translation invariant Gibbs states for the Ising model. Probab. Theory Related Fields , 135(2):153–168, 2006

work page 2006

Show all 41 references

[9]

Bosco, F

G. Bosco, F. Machado, and T. L. Ritchie. Exponential rates of convergence in the ergodic theorem: a constructive approach. Journal of Statistical Physics , 139(3):367–374, 2010. 37

2010
[10]

Burton and J

R. Burton and J. E. Steif. Non-uniqueness of measures of maximal entropy for subshifts of ﬁnite type. Ergodic Theory and Dynamical Systems , 14(2):213–235, 1994

1994
[11]

Burton and J

R. Burton and J. E. Steif. New results on measures of maximal entropy. Israel Journal of Mathematics, 89(1-3):275–300, 1995

1995
[12]

Chayes and K

L. Chayes and K. Shtengel. Lebowitz inequalities for Ashkin–Teller systems. Physica A: Statistical Mechanics and its Applications , 279(1-4):312–323, 2000

2000
[13]

Duminil-Copin

H. Duminil-Copin. Graphical representations of lattice spin models. Lecture Notes (Spartacus-idh, 2015), 2015

2015
[14]

Duminil-Copin

H. Duminil-Copin. Lectures on the Ising and Potts models on the hypercubic lattice. arXiv:1707.00520, 2017

2017 arXiv
[15]

Duminil-Copin, M

H. Duminil-Copin, M. Gagnebin, M. Harel, I. Manolescu, and V. Tassion. Discontinuity of the phase transition for the planar random-cluster and Potts models withq >4. arXiv preprint arXiv:1611.09877, 2016

2016 arXiv
[16]

Duminil-Copin, S

H. Duminil-Copin, S. Goswami, and A. Raouﬁ. Exponential decay of truncated correlations for the Ising model in any dimension for all but the critical temperature. arXiv preprint arXiv:1808.00439 , 2018

2018 arXiv
[17]

Duminil-Copin, V

H. Duminil-Copin, V. Sidoravicius, and V. Tassion. Continuity of the phase transition for planar random-cluster and Potts models with 1≤q≤ 4. Communications in Mathematical Physics, 349(1):47– 107, 2017

2017
[18]

Follmer and S

H. Follmer and S. Orey. Large deviations for the empirical ﬁeld of a Gibbs measure. The Annals of Probability, pages 961–977, 1988

1988
[19]

H.-O. Georgii. Large deviations and maximum entropy principle for interacting random ﬁelds on Zd. The Annals of Probability , pages 1845–1875, 1993

1993
[20]

Georgii, O

H.-O. Georgii, O. H¨ aggstr¨ om, and C. Maes. The random geometry of equilibrium phases. Phase transitions and critical phenomena , 18:1–142, 2001

2001
[21]

Glazman and R

A. Glazman and R. Peled. On the transition between the disordered and antiferroelectric phases of the 6-vertex model. In preparation, 2019

2019
[22]

Grimmett

G. Grimmett. Percolation, volume 321 of Grundlehren der Mathematischen Wissenschaften [Funda- mental Principles of Mathematical Sciences] . Springer-Verlag, Berlin, second edition, 1999

1999
[23]

G. R. Grimmett. The random-cluster model, volume 333. Springer Science & Business Media, 2006

2006
[24]

H¨ aggstr¨ om

O. H¨ aggstr¨ om. On phase transitions for subshifts of ﬁnite type. Israel Journal of Mathematics , 94(1):319–352, 1996

1996
[25]

H¨ aggstr¨ om

O. H¨ aggstr¨ om. Random-cluster representations in the study of phase transitions. 1996

1996
[26]

H¨ aggstr¨ om

O. H¨ aggstr¨ om. Random-cluster analysis of a class of binary lattice gases.Journal of statistical physics , 91(1-2):47–74, 1998. 38

1998
[27]

H¨ aggstr¨ om

O. H¨ aggstr¨ om. Markov random ﬁelds and percolation on general graphs.Advances in Applied Proba- bility, 32(1):39–66, 2000

2000
[28]

Hallberg

P. Hallberg. Gibbs Measures and Phase Transitions in Potts and Beach Models . PhD thesis, Royal Institute of Technology, Stockholm, 2004

2004
[29]

Harel and Y

M. Harel and Y. Spinka. Finitary codings for the random-cluster model and other inﬁnite-range mono- tone models. arXiv preprint arXiv:1808.02333 , 2018

2018 arXiv
[30]

Huang, Y

Y. Huang, Y. Deng, J. L. Jacobsen, and J. Salas. The Hintermann–Merlini–Baxter–Wu and the inﬁnite- coupling-limit Ashkin–Teller models. Nuclear Physics B , 868(2):492–538, 2013

2013
[31]

Lebowitz and G

J. Lebowitz and G. Gallavotti. Phase transitions in binary lattice gases. Journal of Mathematical Physics, 12(7):1129–1133, 1971

1971
[32]

T. M. Liggett. Interacting particle systems, volume 276. Springer Science & Business Media, 2012

2012
[33]

Ornstein and B

D. Ornstein and B. Weiss. Zd-actions and the Ising model. Unpublished, 1977

1977
[34]

Peled and Y

R. Peled and Y. Spinka. A condition for long-range order in discrete spin systems with application to the antiferromagnetic Potts model. arXiv preprint arXiv:1712.03699 , 2017

2017 arXiv
[35]

Peled and Y

R. Peled and Y. Spinka. Lectures on the spin and loop O(n) models. arXiv preprint arXiv:1708.00058, 2017

2017 arXiv
[36]

Peled and Y

R. Peled and Y. Spinka. Rigidity of proper colorings of Zd. arXiv preprint arXiv:1808.03597 , 2018

2018 arXiv
[37]

Pﬁster and Y

C.-E. Pﬁster and Y. Velenik. Random-cluster representation of the Ashkin-Teller model. Journal of statistical physics, 88(5-6):1295–1331, 1997

1997
[38]

Pisztora

A. Pisztora. Surface order large deviations for Ising, Potts and percolation models. Probab. Theory Related Fields, 104(4):427–466, 1996

1996
[39]

A. Raouﬁ. Translation-invariant gibbs states of Ising model: general setting. arXiv preprint arXiv:1710.07608, 2017

2017 arXiv
[40]

Ray and Y

G. Ray and Y. Spinka. A short proof of the discontinuity of phase transition in the planar random- cluster model with q >4. arXiv preprint arXiv:1904.10557 , 2019

1904 arXiv
[41]

Y. Spinka. Finitary codings for spatial mixing Markov random ﬁelds. arXiv preprint arXiv:1803.10578, 2018. Emails: gourabray@uvic.ca, yinon@math.ubc.ca 39

2018 arXiv

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

S. Adams. Følner independence and the amenable Ising model. Ergodic Theory and Dynamical Systems, 12(4):633–657, 1992

work page 1992

[2] [2]

Aizenman, H

M. Aizenman, H. Duminil-Copin, and V. Sidoravicius. Random currents and continuity of Ising models spontaneous magnetization. Communications in Mathematical Physics , 334(2):719–742, 2015

work page 2015

[3] [3]

Emergent Planarity in two-dimensional Ising Models with finite-range Interactions

M. Aizenman, H. Duminil-Copin, V. Tassion, and S. Warzel. Emergent planarity in two-dimensional ising models with ﬁnite-range interactions. Inventiones, to appear, 2018. arXiv:1801.04960

work page Pith review arXiv 2018

[4] [4]

R. J. Baxter, S. B. Kelland, and F. Y. Wu. Equivalence of the Potts model or Whitney polynomial with an ice-type model. Journal of Physics A: Mathematical and General , 9(3):397, 1976

work page 1976

[5] [5]

J. v. d. Berg and C. Maes. Disagreement percolation in the study of Markov ﬁelds. The Annals of Probability, 22(2):749–763, 1994

work page 1994

[6] [6]

J. v. d. Berg and J. E. Steif. On the existence and nonexistence of ﬁnitary codings for a class of random ﬁelds. Annals of probability, pages 1501–1522, 1999

work page 1999

[7] [7]

Bodineau

T. Bodineau. Slab percolation for the Ising model. Probab. Theory Related Fields, 132(1):83–118, 2005

work page 2005

[8] [8]

Bodineau

T. Bodineau. Translation invariant Gibbs states for the Ising model. Probab. Theory Related Fields , 135(2):153–168, 2006

work page 2006