Pith. sign in

REVIEW 3 major objections 4 minor 49 references

Dimension jump at the uniqueness threshold for percolation in $\infty+d$ dimensions

T0 review · 3 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read The paper proves that for percolation on tree-with-lattice graphs, the boundary dimension of an infinite cluster is at most 1/2 at the uniqueness threshold and jumps to 1 immediately above it.

desk verdict A genuinely new dimension jump for unoriented percolation on tree products, but the exact dimension formula rests on an omitted lower bound that must be supplied before the paper is complete. read the letter →

arxiv 2412.15895 v1 pith:VI5HWAL2 submitted 2024-12-20 math.PR

classification math.PR MSC 60K3582B4360J80
keywords percolationuniquenessthresholdHausdorffdimensionnonunimodulartransitivegraphslamplightergraphtreeproductstiltedsusceptibilitybranchingrandomwalk
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

This paper studies Bernoulli percolation on the product of a regular tree with an infinite amenable graph, and on the lamplighter graph over the tree. It establishes that as the percolation parameter crosses the uniqueness threshold $p_u$, the set of accumulation points of an infinite cluster in the boundary of the tree changes dimension discontinuously: at most $1/2$ below and at $p_u$, and full dimension $1$ above $p_u$. The dimension is identified almost surely with $1-\beta_p^*$, where $\beta_p^*$ is an exponent extracted from slab-to-fiber connection probabilities. Along the way the paper proves that four central thresholds—$p_u$, the $L^2$ boundedness threshold, the tiltability threshold, and the heaviness threshold—coincide for these graphs, giving the first nonamenable examples known to have $p_{2\to2}=p_u$. If correct, this makes the percolation transition on these graphs resemble the branching random walk recurrence transition in a quantitative, geometric way.

What carries the argument

The proof is carried by the exponent $\beta_p^*$, defined from the supermultiplicative sequence $P_p(n)=P_p(o\leftrightarrow[x])$ for a fiber at height $-n$, with $\beta_p^*=-\lim_n\log P_p(n)/(n\log(k-1))$; a standard supermultiplicativity argument gives the limit, and Lemma 2.4 ties it to geometry: the boundary limit set has Hausdorff dimension $1-\beta_p^*$ on the event that the cluster is infinite. The nonunimodular automorphism group fixing an end of the tree supplies the tilted mass-transport principle, with modular function $\Delta(x,y)=(k-1)^{h(x,y)}$, used to define tilted susceptibilities and the thresholds $p_t$ and $p_h$. Backscattering lemmas show $\beta_p^*\ge 1/2$ below $p_u$ by constructing paths that descend deep into the tree and return to the root fiber; for products with a lattice this uses an embedded branching random walk on the amenable factor and its spectral-radius behavior. Finally, a bridge-decomposition argument in the style used for self-avoiding walks bounds the tilted susceptibility $\chi_{p,1/2}$ by a product of slab generating functions, proving $p_t=p_u$ and hence the threshold equalities.

What would settle it

Find a $p<p_u$ and an amenable $H$ for which the boundary limit set of an infinite cluster has Hausdorff dimension strictly larger than $1/2$; Theorem 1.1 says this cannot happen. A concrete route is to run percolation on $T\times\mathbb{Z}$ at a numerically sub-$p_u$ parameter, measure the box-counting or Hausdorff dimension of the cluster's accumulation set on $\partial T$, and compare it with $1-\beta_p^*$ estimated from the decay of $P_p(n)$; any systematic gap between the measured dimension and $1-\beta_p^*$ below $p_u$ would falsify Lemma 2.4.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1.1: for percolation on $T\times H$ with $H$ an infinite amenable Cayley graph, or on the lamplighter graph $LL(T)$, conditional on the cluster of the root being infinite, the Hausdorff dimension $\delta_H(\Lambda)$ of the set of limit points in $\partial T$ of the projected cluster is $P_p$-almost surely equal to a constant $\delta_H(p)$. This constant is $1-\beta_p^*$ for $p_c<p\le p_u$, with $\delta_H(p_u)\le 1/2$, and is $1$ for every $p>p_u$; moreover $p\mapsto\delta_H(p)$ is continuous on $(p_c,p_u]$. Theorem 1.4 adds $p_u=p_{2\to2}=p_t=p_h$ for these graphs. The statement strengthens the known non-uniqueness at $p_u$: instead of merely many infinite clusters, each cluster's boundary footprint is geometrically small (dimension at most $1/2$) at and below the threshold and full (dimension $1$) above it.

Load-bearing premise

The load-bearing premise is the matching lower bound in Lemma 2.4, whose proof is only sketched with the sentence 'we omit further details here' and deferred to the techniques of [18,33]; the paper also relies on the companion result [26] for the exponential tail, so the equality $\delta_H(p)=1-\beta_p^*$ stands or falls with those deferred ingredients.

Editorial extensions

If this is right

  • For every $p\le p_u$, any infinite cluster's projected cluster accumulates on a boundary set of Hausdorff dimension at most $1/2$, so infinite clusters occupy at most a square-root fraction of the boundary.
  • For every $p>p_u$, the projected cluster is dense in the boundary and its accumulation set has full dimension $1$, matching the trivial upper bound.
  • The equality $p_{2\to2}=p_u$ provides the first nonamenable transitive graphs for which the $L^2$ boundedness threshold sits at the uniqueness threshold rather than strictly below it.
  • The coincidence $p_u=p_t=p_h$ integrates the uniqueness threshold with the nonunimodular tiltability and heaviness thresholds for these graphs.
  • The dimension jump mirrors the branching random walk recurrence transition, giving a quantitative sense in which percolation on these graphs behaves like branching random walk up to $p_u$.

Reading between the lines

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

  • Beyond the paper, the same mechanism should extend to products of trees with any amenable lamp group or with switch-walk-switch Cayley graphs, since the proof's key input is the amenability of fibers rather than the specific lattice structure.
  • The equality $p_{2\to2}=p_u$ suggests a testable dichotomy: on graphs with $p_c<p_u$, the position of $p_{2\to2}$ relative to $p_u$ may encode whether infinite clusters below $p_u$ are tree-like or dense, and simulations on hyperbolic tessellations versus $T\times\mathbb{Z}$ could calibrate this.
  • Because $\beta_p^*$ is a supermultiplicative rate, one could estimate it from finite-slab simulations and use the predicted dimension jump as a numerical locator for $p_u$ on infinite-dimensional graphs.
  • The omitted lower-bound half of Lemma 2.4 is the part most worth checking: if the embedded branching-process construction cannot be made to give the matching dimension, the theorem would still hold as an upper bound but the exact dimension formula would be lost.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper studies Bernoulli percolation on the product T×H of a k-regular tree T with an infinite amenable Cayley graph H, and on the lamplighter graph LL(T). It claims that on the event of an infinite cluster, the Hausdorff dimension of the cluster's limit set in the tree boundary is at most 1/2 below the uniqueness threshold p_u and jumps to 1 above p_u. It also claims equality of several critical thresholds: the uniqueness threshold p_u, the L2-boundedness threshold p_{2→2}, the tiltability threshold p_t, and the heaviness threshold p_h. The proofs use nonunimodular percolation techniques, a supermultiplicative sequence (P_p(n)) with rate β*_p, and a Hammersley-Welsh argument to show χ_{p,1/2}<∞ for p<p_u.

Significance. If the claims hold, this constitutes the first proof of a dimension jump for unoriented percolation and the first nonamenable examples with p_{2→2}=p_u. The paper is well-structured and contains detailed arguments in Sections 2.2-2.4, including the backscattering lemmas and the Hammersley-Welsh decomposition. However, the central dimension formula rests on a lower bound in Lemma 2.4 that is explicitly omitted, and two key lemmas are imported from the same-authors' companion preprint [26]. The lower-bound gap is fillable in principle, and the companion-paper dependence is a matter of verification, so the work has substantial potential but is not yet complete as a standalone paper.

major comments (3)
  1. [Section 2.1, Lemma 2.4] The lower bound δ_H(Λ) ≥ 1−β*_p for p_c<p≤p_u is asserted but not proved. The text states that the matching lower bound follows by a 'very similar' argument to [18,33] and 'we omit further details here,' deferring to Lemma 2.8. However, Lemma 2.8 constructs embedded branching processes that control the probability Q_p(n) of a single first-arrival connection to a fiber; it does not construct a boundary measure or a Frostman-type lower bound for the limit set. Without this lower bound, Theorem 1.1's exact formula δ_H(p)=1−β*_p, and the claimed continuous increase from 0 to δ_H(p_u), are not established. Only the upper bound δ_H≤1−β*_p≤1/2 below p_u would follow. This is a load-bearing gap in the proof of the main theorem.
  2. [Sections 2.1–2.2, Lemmas 2.1 and 2.6] Lemmas 2.1 and 2.6 are stated without proof and are attributed to the companion preprint [26]. Lemma 2.1 is used to identify p_u with the threshold for finite fiber intersections, and Lemma 2.6 provides the exponential tail estimate that is essential in Lemma 2.5 and in the proofs of Lemmas 2.7 and 2.9. Since [26] is a same-author preprint that is not yet a published reference, the present paper's central claims cannot be verified independently of [26]. The authors should either provide full proofs of these statements in an appendix or ensure that [26] is accepted and publicly available in a verifiable form before the paper is accepted.
  3. [Section 2.3, Lemma 2.9] In the proof of Lemma 2.9, the final step asserts that if the branching process Y^{h,g} survives, then 'in either case, the two clusters will become a.s. equal when we increase p to any p'>p in the standard monotone coupling.' This step is not proved. To justify it, one must argue that the infinitely many levels at which the clusters come within distance d(h,g) give rise to infinitely many edge-disjoint opportunities to connect the clusters at p', so that by Borel-Cantelli the merge occurs almost surely. This is plausible and likely fixable, but as written the merging argument is a genuine gap in the proof that β*_p≥1/2 for T×H.
minor comments (4)
  1. [Section 2.1, Lemma 2.4 (statement)] The statement contains the typo 'for each p_c < p c≤ p_u'; it should be 'for each p_c < p ≤ p_u'.
  2. [Section 2.2, Lemma 2.5] The expression 'log e/Pp(n)' is ambiguous. It should be written as log(e/Pp(n)) to mean the natural logarithm of the ratio e divided by Pp(n).
  3. [Section 2.4, proof of Lemma 2.10] There is a notation inconsistency: the sum is written over s∈S_{n,k}, while the set was previously defined as S_{l,n}. The index set should be S_{l,n} (or S_{n,l}), and the quantities should be labeled consistently.
  4. [Section 2.1, proof of Lemma 2.2] The lower bound 'by the Harris-FKG inequality' is not precisely a Harris-FKG statement; the argument is a union bound conditional on the event that at least one vertex in the fiber is connected, combined with independence of the extension edges. The sentence could be rephrased to avoid implying an FKG correlation inequality where none is needed.

Circularity Check

2 steps flagged · score 4.0 of 10

Load-bearing estimates imported from same-authors companion preprint; exact dimension formula rests on an omitted lower-bound proof, but no definitional circularity.

  1. self citation load bearing [Section 2.2, Lemma 2.6 and its use in Lemma 2.5]
    "We will deduce Lemma 2.5 from the following lemma, which is an immediate consequence of the main results of our companion paper [26]. Lemma 2.6. If p < p_u then sup_x P_p(|K_o ∩ [x]| ≥ n) decays exponentially fast in n."

    The exponential tail and finiteness of E_p|K_o ∩ [x]| below p_u are imported from the same-authors preprint [26]. This estimate is load-bearing: Lemma 2.5 uses it to compare P_p(n) and E_p(n) up to subexponential factors, and that comparison is used in the proof of Theorem 1.4 and in the continuity argument for Theorem 1.1. The paper does not prove the estimate here, so a central part of the derivation chain is supported by a self-citation to a companion preprint rather than by independent evidence.

  2. self citation load bearing [Section 2.1, Lemma 2.1]
    "Lemma 2.1. Every cluster has finite intersection with every fiber [x] a.s. if and only if p ≤ p_u. Proof of lemma 2.1. Since the fiber [o] is an amenable normal subgroup of a group for which G is a Cayley graph, it follows from [26, Theorem 1.9 and Proposition 1.9] that the intersection |K_o ∩ [o]| is almost surely finite if and only if there is not a unique infinite cluster in G."

    This lemma gives the paper's operative characterization of the uniqueness threshold p_u, namely finiteness of the cluster's intersection with every fiber. The proof is entirely delegated to results in [26], another paper by the same two authors. The lemma is then used in the backscattering arguments (Lemmas 2.7 and 2.9) to conclude p ≥ p_u when the cluster can meet a fiber infinitely often, and in the proof of Theorem 1.4. This is load-bearing self-citation, though the cited companion result is a stated theorem with its own assumptions and is not a definitional restatement of the target claims.

full rationale

The central claims of the paper -- the Hausdorff dimension jump at p_u, the equality p_u = p_{2->2} = p_t = p_h, and the continuous increase of δ_H(p) below p_u -- are genuine new statements and are not obtained by renaming known quantities or by fitting parameters. The proofs contain substantial independent content: the supermultiplicativity argument, the embedded branching-process constructions in Lemmas 2.8 and 2.9, the lamplighter backscattering inequality, and the Hammersley-Welsh bound for the tilted susceptibility are all derived in the paper and do not reduce by construction to the theorems they prove. The external result of Schonmann is used as independent grounding for non-uniqueness at p_u. However, the derivation chain is not self-contained: two load-bearing facts, Lemma 2.1 and Lemma 2.6, are imported from the same-authors companion preprint [26], and the proof of Theorem 1.4 explicitly relies on Lemma 2.6. In addition, the exact equality δ_H(Λ) = 1 - β*_p in Theorem 1.1 depends on the lower-bound half of Lemma 2.4, which the paper does not prove: the text says 'we omit further details here' and refers to Lemma 2.8, but Lemma 2.8 only controls the point-to-fiber probability Q_p(n) and does not by itself supply the Frostman-type lower bound on the limit set. This is a proof gap rather than a circularity, because no equation in Lemma 2.4 is defined in terms of the Hausdorff dimension it claims to compute. Weighing all of this, the paper has significant self-citation and an omitted load-bearing proof, but the central claims still have independent mathematical content; the score is therefore 4 rather than higher.

Assumptions & free parameters 0 free parameters · 8 assumptions · 0 invented entities

The proof has no fitted constants or invented entities. It relies on standard percolation theorems and on the nonunimodular framework of [23]; the companion paper [26] supplies the key relative sharpness estimates. The main new content is the backscattering and Hammersley-Welsh machinery.

assumptions (8)
  • standard math For transitive graphs, the number of infinite clusters is 0, 1, or infinity and changes only at p_c and p_u, per Burton-Keane and Haggstrom-Peres-Schonmann.
    Used throughout to define the phase diagram and to justify Schonmann's non-uniqueness at p_u.
  • standard math Lyons-Schramm indistinguishability theorem: all infinite clusters have the same limit set dimension a.s.
    Invoked in Lemma 2.4 to make the Hausdorff dimension of the limit set non-random.
  • standard math Kesten's theorem on amenable groups: return probabilities of a random walk on an amenable Cayley graph decay subexponentially.
    Used in Lemma 2.8 to control the branching random walk on the amenable factor H.
  • standard math Harris-FKG and BK inequalities for Bernoulli percolation.
    Used in supermultiplicativity, backscattering, and Hammersley-Welsh estimates throughout Section 2.
  • standard math Fekete's lemma and basic properties of supermultiplicative and submultiplicative sequences.
    Defines β*_p, α_p, and β_p and gives their convergence and continuity properties.
  • domain assumption Hutchcroft [23]: nonunimodular percolation framework, tilted susceptibility, the inequality p_t≤p_{2->2}≤p_u, and continuity results for α_p and β_p.
    The whole proof is set in the nonunimodular framework of [23]; these results are cited but not reproved.
  • domain assumption Hutchcroft-Pan [26]: relative sharpness of the phase transition and the characterization of p_u via finite intersections with fibers, Lemmas 2.1 and 2.6.
    Key inputs from a companion preprint by the same authors; the present paper is not self-contained and inherits any error there.
  • ad hoc to paper Lower bound in Lemma 2.4: δ_H ≥ 1-β*_p via embedded Galton-Watson processes and the Hawkes-Lyons theorem.
    Asserted without proof in this paper with 'we omit further details here'; essential to the exact dimension formula in Theorem 1.1.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Dimension jump at the uniqueness threshold for percolation in $\infty+d$ dimensions." pith.science (2026). https://pith.science/paper/VI5HWAL2

@misc{pith2026241215895,
  author       = {Pith},
  title        = {Pith review of: Dimension jump at the uniqueness threshold for percolation in $\infty+d$ dimensions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VI5HWAL2}},
  note         = {Machine review of arXiv:2412.15895}
}
abstract

Consider percolation on $T\times \mathbb{Z}^d$, the product of a regular tree of degree $k\geq 3$ with the hypercubic lattice $\mathbb{Z}^d$. It is known that this graph has $0<p_c<p_u<1$, so that there are non-trivial regimes in which percolation has $0$, $\infty$, and $1$ infinite clusters a.s., and it was proven by Schonmann (1999) that there are infinitely many infinite clusters a.s. at the uniqueness threshold $p=p_u$. We strengthen this result by showing that the Hausdorff dimension of the set of accumulation points of each infinite cluster in the boundary of the tree has a jump discontinuity from at most $1/2$ to $1$ at the uniqueness threshold $p_u$. We also prove that various other critical thresholds including the $L^2$ boundedness threshold $p_{2\to 2}$ coincide with $p_u$ for such products, which are the first nonamenable examples proven to have this property. All our results apply more generally to products of trees with arbitrary infinite amenable Cayley graphs and to the lamplighter on the tree.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 47 canonical work pages

  1. [26]

    Percolation at the uniqueness threshold via subgroup relativization

    T. Hutchcroft and M. Pan. Percolation at the uniqueness thre shold via subgroup relativization. arXiv preprint arXiv:2409.12283, 2024

  2. [23]

    Hutchcroft

    T. Hutchcroft. Nonuniqueness and mean-field criticality for pe rcolation on nonunimodular transitive graphs. J. Amer. Math. Soc. , 33(4):1101–1165, 2020

  3. [1]

    Aizenman and C

    M. Aizenman and C. M. Newman. Tree graph inequalities and critical behavior in percolation models. J. Statist. Phys. , 36(1-2):107–143, 1984

  4. [2]

    D. J. Barsky and M. Aizenman. Percolation critical exponents un der the triangle condition. Ann. Probab., 19(4):1520–1536, 1991

  5. [3]

    Benjamini and O

    I. Benjamini and O. Schramm. Percolation beyond Zd, many questions and a few answers. Electron. Comm. Probab., 1:no. 8, 71–82, 1996

  6. [4]

    Benjamini and O

    I. Benjamini and O. Schramm. Percolation in the hyperbolic plane. J. Amer. Math. Soc. , 14(2):487–507, 2001

  7. [5]

    R. M. Burton and M. Keane. Density and uniqueness in percolation . Communications in mathematical physics, 121(3):501–505, 1989

  8. [6]

    Candellero, L

    E. Candellero, L. A. Gilch, and S. M¨ uller. Branching random walks o n free products of groups. Proceedings of the London Mathematical Society , 104(6):1085–1120, 2012

Show all 49 references
  1. [7]

    Duminil-Copin and R

    H. Duminil-Copin and R. Panis. An alternative approach for the mea n-field behaviour of spread-out Bernoulli percolation in dimensions d > 6. arXiv preprint arXiv:2410.03647 , 2024

  2. [8]

    Dussaule, L

    M. Dussaule, L. Wang, and W. Yang. Branching random walks on re latively hyperbolic groups. arXiv preprint arXiv:2211.07213, 2022

  3. [9]

    Fitzner and R

    R. Fitzner and R. van der Hofstad. Mean-field behavior for near est-neighbor percolation in d > 10. Electron. J. Probab., 22:Paper No. 43, 65, 2017

  4. [10]

    Gaboriau and R

    D. Gaboriau and R. Tucker-Drob. Approximations of standard equivalence relations and Bernoulli perco- lation at pu. Comptes Rendus. Math´ ematique, 354(11):1114–1118, 2016

  5. [11]

    Gantert and S

    N. Gantert and S. Mueller. The critical branching Markov chain is transient. Markov Process. Related Fields, 2005

  6. [12]

    Grimmett

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

  7. [13]

    G. R. Grimmett and C. M. Newman. Percolation in ∞ + 1 dimensions. In Disorder in physical systems , Oxford Sci. Publ., pages 167–190. Oxford Univ. Press, New York, 1 990

  8. [14]

    H¨ aggstr¨ om, Y

    O. H¨ aggstr¨ om, Y. Peres, and R. H. Schonmann. Percolation on transitive graphs as a coalescent process: relentless merging followed by simultaneous uniqueness. In Perplexing problems in probability , volume 44 of Progr. Probab., pages 69–90. Birkh¨ auser Boston, Boston, MA, 1999

  9. [15]

    J. M. Hammersley and D. J. A. Welsh. Further results on the rat e of convergence to the connective constant of the hypercubical lattice. Quart. J. Math. Oxford Ser. (2) , 13:108–110, 1962

  10. [16]

    Hara and G

    T. Hara and G. Slade. Mean-field critical behaviour for percolat ion in high dimensions. Comm. Math. Phys. , 128(2):333–391, 1990

  11. [17]

    J. Hawkes. Trees generated by a simple branching process. Journal of the London Mathematical Society , s2-24(2):373–384, 1981

  12. [18]

    Hueter and S

    I. Hueter and S. P. Lalley. Anisotropic branching random walks o n homogeneous trees. Annals of Probability, 29(4):1563–1590, 2000

  13. [19]

    Hutchcroft

    T. Hutchcroft. Percolation on hyperbolic graphs. Geometric and Functional Analysis , 29(3):766–810, Jun 2019

  14. [20]

    Hutchcroft

    T. Hutchcroft. Self-avoiding walk on nonunimodular transitive g raphs. Ann. Probab. , 47(5):2801–2829, 2019

  15. [21]

    Hutchcroft

    T. Hutchcroft. The L2 boundedness condition in nonamenable percolation. Electron. J. Probab. , 25:Paper No. 127, 27, 2020. 18 TOM HUTCHCROFT AND MINGHAO PAN

  16. [22]

    Hutchcroft

    T. Hutchcroft. Non-intersection of transient branching ran dom walks. Probability Theory and Related Fields, 178(1):1–23, 2020

  17. [24]

    Hutchcroft

    T. Hutchcroft. Slightly supercritical percolation on non-amen able graphs I: The distribution of finite clus- ters. Proceedings of the London Mathematical Society , 125(4):968–1013, 2022

  18. [25]

    Hutchcroft

    T. Hutchcroft. Slightly supercritical percolation on nonamena ble graphs II: growth and isoperimetry of infinite clusters. Probability Theory and Related Fields , 188(1):549–582, 2024

  19. [27]

    Hutchcroft and M

    T. Hutchcroft and M. Tointon. Non-triviality of the phase tran sition for percolation on finite transitive graphs. Journal of the European Mathematical Society , 2024

  20. [28]

    F. I. Karpelevich, E. Pechersky, and Y. M. Suhov. A phase tra nsition for hyperbolic branching processes. Communications in mathematical physics , 195(3):627–642, 1998

  21. [29]

    H. Kesten. Full banach mean values on countable groups. Mathematica Scandinavica, pages 146–156, 1959

  22. [30]

    S. P. Lalley. Percolation clusters in hyperbolic tessellations. Geometric And Functional Analysis , 11:971–1030, 2001

  23. [31]

    S. P. Lalley. The weak/strong survival transition on trees and nonamenable graphs. In International congress of mathematicians , volume 3, pages 637–647, 2006

  24. [32]

    S. P. Lalley and T. Sellke. Hyperbolic branching Brownian motion. Probability theory and related fields , 108:171–192, 1997

  25. [33]

    S. P. Lalley and T. Sellke. Limit set of a weakly supercritical conta ct process on a homogeneous tree. Annals of Probability, pages 644–657, 1998

  26. [34]

    T. Liggett. Branching random walks and contact processes on homogeneous trees. Probability Theory and Related Fields, 106:495–519, 1996

  27. [35]

    T. M. Liggett. Branching random walks and contact processes on homogeneous trees. Probability theory and related fields , 106:495–519, 1996

  28. [36]

    R. Lyons. Random walks and percolation on trees. The annals of Probability , 18(3):931–958, 1990

  29. [37]

    R. Lyons. Fixed price of groups and percolation. Ergodic Theory Dynam. Systems , 33(1):183–185, 2013

  30. [38]

    Lyons and Y

    R. Lyons and Y. Peres. Probability on Trees and Networks , volume 42 of Cambridge Series in Statistical and Probabilistic Mathematics . Cambridge University Press, New York, 2016

  31. [39]

    Lyons and O

    R. Lyons and O. Schramm. Indistinguishability of percolation clus ters. Ann. Probab. , 27(4):1809–1836, 1999

  32. [40]

    Nachmias and Y

    A. Nachmias and Y. Peres. Non-amenable Cayley graphs of high g irth have pc < p u and mean-field expo- nents. Electron. Commun. Probab. , 17:no. 57, 8, 2012

  33. [41]

    Pak and T

    I. Pak and T. Smirnova-Nagnibeda. On non-uniqueness of perc olation on nonamenable Cayley graphs. C. R. Acad. Sci. Paris S´ er. I Math. , 330(6):495–500, 2000

  34. [42]

    Y. Peres. Percolation on nonamenable products at the uniquen ess threshold. Ann. Inst. H. Poincar´ e Probab. Statist., 36(3):395–406, 2000

  35. [43]

    R. H. Schonmann. Percolation in ∞ + 1 dimensions at the uniqueness threshold. Perplexing Problems in Probability: Festschrift in Honor of Harry Kesten , pages 53–67, 1999

  36. [44]

    R. H. Schonmann. Stability of infinite clusters in supercritical pe rcolation. Probab. Theory Related Fields , 113(2):287–300, 1999

  37. [45]

    R. H. Schonmann. Multiplicity of phase transitions and mean-field criticality on highly non-amenable graphs. Comm. Math. Phys. , 219(2):271–322, 2001. DIMENSION JUMP AT pu FOR PERCOLATION IN ∞ + d DIMENSIONS 19

  38. [46]

    R. H. Schonmann. Mean-field criticality for percolation on planar non-amenable graphs. Comm. Math. Phys., 225(3):453–463, 2002

  39. [47]

    Sidoravicius, L

    V. Sidoravicius, L. Wang, and K. Xiang. Limit set of branching ran dom walks on hyperbolic groups. Communications on Pure and Applied Mathematics , 76(10):2765–2803, 2023

  40. [48]

    A. Tim´ ar. Percolation on nonunimodular transitive graphs. Ann. Probab., 34(6):2344–2364, 2006

  41. [49]

    W. Woess. Random walks on infinite graphs and groups , volume 138 of Cambridge Tracts in Mathematics . Cambridge University Press, Cambridge, 2000

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.