Pith. sign in

REVIEW 3 major objections 4 minor 2 cited by

Rapid phase ordering for Ising and Potts dynamics on random regular graphs

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The paper proves that low-temperature Ising and Potts Glauber dynamics on random d-regular graphs, started with a small plurality bias toward one state, reach the phase-restricted stationary distribution in O(log n) time and stay close to…

desk verdict A substantial new proof that biased Ising and Potts dynamics on random d-regular graphs reach the majority phase in O(log n) time; the rigid-dynamics machinery is genuinely new, and the main residual risk is the imported stationary-phase concentration estimate. read the letter →

arxiv 2505.15783 v1 pith:CF2NFCO2 submitted 2025-05-21 math.PR math-phmath.MP

Reza Gheissari , Allan Sly , Youngtak Sohn This is my paper · ORCID
classification math.PRmath-phmath.MP MSC 60K3582B2082C2005C80
keywords IsingmodelPottsGlauberdynamicsrandomregulargraphsphaseorderingmixingtimemetastabilitytotalvariationdistance
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's goal is to show that the only obstacle to fast equilibration of low-temperature Ising and Potts dynamics on random d-regular graphs is the bottleneck between the q dominant phases, and that once the dynamics enters a phase, equilibration is fast. It proves that starting with even a small excess of vertices in state 1, an excess that can shrink to zero as the degree d grows, the Glauber dynamics converges in O(log n) time to the stationary distribution conditioned on state 1 having plurality, and stays close to it for exponential time. This turns a previously open question, quasi-equilibration from non-monochromatic initializations, into a theorem for every d ≥ 7. The interest is that the model is deep in the low-temperature regime where local correlations are strong and standard mixing tools fail, so the proof needs a genuinely new way to confine negative information in spacetime.

What carries the argument

The argument is carried by a spacetime cluster analysis in the grand coupling of Glauber chains. Minus spacetime clusters confine negative information, and the legacy region is the union of clusters touching the initial minus set; once it dies out, the biased chain is perfectly coupled to the all-plus chain. To get exponential tails on cluster sizes, the paper introduces the rigid dynamics, a coupled non-Markovian variant that rejects flips from −1 to +1 at trifurcation points, meaning vertices whose minus region touches at least three subtrees of their local neighborhood. A delicate temporal recursion on the probability mass function of spacetime cluster sizes, with threshold ψ_k = $k^{{−2}}$$d^{{−1000−100k}}$, shows those sizes are subcritical, so the legacy region has negative drift and dies out in O(log n) time.

What would settle it

Run the grand-coupling simulation on a fixed random 7-regular graph of moderate size, at β = C log 7 / 7 with C a large constant, starting from an initialization with ε n pluses for ε around 1/log 7: the theorem predicts the biased and all-plus chains couple by time C log n with failure probability at most $n^{{−10}}$, so observing coupling times that grow faster than logarithmic, or minus spacetime clusters of size comparable to R = (1/4) log_d n, would refute the central claim.

Watch

Extended reading notes

Core claim

For every degree d ≥ 7 and every number q ≥ 2 of Potts colors, at inverse temperatures β above a constant times log d / d, the following holds with probability 1 − o(1) over the random graph: if the initial configuration has m(X0) ≥ ε with ε > ε0(d), where ε0 ≍ 1/log d for the Ising case and ε0 ≍ max{$d^{{−1/2}}$, (βp d)^{−1}} for Potts, then the total-variation distance between the law of the dynamics and π conditioned on the state-1 plurality phase is at most $n^{{−10}}$ for every time between C log n and $e^{{n/C}}$. Equivalently, the chain quasi-equilibrates to the metastable phase in optimal logarithmic time, even though the unconditional mixing time is exponential. The authors establish this by showing that the legacy region of spacetime negative information inherited from the initialization dies out in O(log n) time.

Load-bearing premise

The proof imports a concentration estimate on the target conditional measure: with high probability over the graph, the phase-restricted Gibbs measure π1 puts at least 1 − $e^{{−cn}}$ of its mass on configurations with a large majority of vertices in state 1, and if that external estimate failed, the all-phase chain used in the triangle inequality would not be close to π1 and the total-variation bound would not follow.

Editorial extensions

If this is right

  • For every d ≥ 7 and β at least a large constant times log d / d, Ising and q-state Potts dynamics from biased starts reach the phase-restricted stationary distribution within O(log n) time and remain there up to time e^{n/C}.
  • The required initial bias tends to zero as d → ∞: ε0(d) ≍ 1/log d for the Ising case, and it can be improved to ≍ d^{−1/2} at a slightly larger temperature threshold.
  • The set of two-spin dynamics for which the legacy-extinction argument works is described by two explicit conditions, monotonicity and a low-temperature flip probability, so the result applies, for example, to the noisy majority model with sufficiently small noise.
  • If the initial bias is at least 1 − γ0 for a universal constant γ0, the logarithmic-time bound holds uniformly in d ≥ 7; the small-bias results follow from a separate magnetization-drift argument at large d.
  • The exponential upper bound on the time window is necessary, since after exponential time the chain fully mixes and gives probability 1/2 to the opposite phase.

Reading between the lines

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

  • Beyond the stated theorems, the proof's reliance on local treelikeness up to radius (1/4) log_d n suggests the d ≥ 7 barrier comes from short cycles that are all-minus by chance; extending to d ≥ 3 would require handling those rare local obstructions, as the authors remark.
  • Because only monotonicity and a lower bound on plus-flip probabilities are used, the rigid-dynamics subcriticality method looks portable to other ferromagnetic models, including non-reversible ones such as noisy majority dynamics, for which spectral and functional-analytic tools are limited.
  • A quantitative prediction left implicit is strengthened by the paper's comment that C can be taken to be 1 + o_d(1): for large d, quasi-equilibration from a biased start should occur in (1 + o(1)) log n time.
  • For low-temperature sampling on random regular graphs, the result implies that a phase-restricted or biased initialization is enough to avoid the exponential bottleneck, so random restarts inside a phase should give provably short burn-in in practice.
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 Ising and q-state Potts Glauber dynamics on random d-regular graphs at low temperatures β ≍ log d / d. It proves that from initializations with a small bias toward one phase, the dynamics reaches total-variation distance n^{-10} from the conditional stationary measure restricted to that phase within O(log n) time, and stays there for exponentially long times. For Ising the required initial bias can be taken ε0 ≍ 1/log d, and for Potts ε0 ≍ max{d^{-1/2}, (βp d)^{-1}}. The proof introduces a non-Markovian 'rigid dynamics' whose minus spacetime clusters are controlled by a temporal recursion on explicit exponential-tailed mass functions, and then analyzes the negative drift of the 'legacy region' of minus clusters originating from the initialization. The final step couples the Potts dynamics to a two-spin dominating process and uses a triangle inequality with a restricted chain initialized from the stationary plus phase.

Significance. If the proof is correct, this is a substantial advance: it gives optimal O(log n) quasi-equilibration from biased, non-monochromatic initializations in a setting where worst-case mixing is exponentially slow, and it does so uniformly over a growing degree d with bias tending to zero. The rigid-dynamics construction and the conditional spacetime-cluster recursion are new and likely to be influential for other low-temperature ferromagnetic models. The paper is largely self-contained modulo standard external results (Friedman's theorem and the cited concentration properties of the stationary measure). The explicit thresholds ψ_k are constructed rather than fitted, and the proof does not rely on circular use of the conclusion.

major comments (3)
  1. [Sec. 4.2, Lemma 4.6] The coupling used to prove the stochastic domination of the conditioned dynamics by the all-plus dynamics is not valid as written. In the case λ_v(r) = c_{v,s+r}, the paper flips X_{s+r}(v) with probability 1 and X_{s+r}^+(v) with probability 1 - (1/c) (e^{c - c^+} - 1). This quantity is not a probability: for c > c^+, it can be negative, and it does not reduce to the standard thinning probability c^+/c. Since Corollary 4.7 and the subsequent bounds on the expected sizes of non-legacy minus regions depend on this domination, the proof needs to be corrected, for example by replacing the displayed expression with the standard rate-c^+/c thinning rule.
  2. [Sec. 3.2, Lemma 3.9, case (B)] The displayed computation of the probability that a non-trifurcation vertex is in the event {|R_{t-}(u)| = k, |C_{t-}(u)| = ℓ} has the wrong sign. The equality should read ρ_t(u; k, ℓ) - ρ_t(u; k+1, ℓ), not ρ_t(u; k+1, ℓ) - ρ_t(u; k, ℓ). With the printed sign, the bound eP ≤ -ψ_k ψ_ℓ / 3 is impossible for a probability, and the negative contribution from the shrinking term in (3.11) is not established. Reversing the sign gives the needed lower bound eP ≥ ψ_k ψ_ℓ / 3 once ψ_{k+1} ≤ d^{-100} ψ_k is used. This is a local sign error, but it sits in the core temporal recursion and must be fixed.
  3. [Sec. 4.4, Eq. (4.6)] The total-variation bound in Theorem 4.1 depends on the imported concentration estimate π1(σ : m(σ) ≥ 1 - γ0) ≥ 1 - e^{-cn} at βp > C log q / d. The proof does not state the precise external theorem or the dependence of C and c(C). This estimate is load-bearing: it is used both to compare Y_t^{π1} with the restricted chain and to control the restricted chain's hitting probability. If the cited results from [13] and [2,12] only give such concentration at larger β or with weaker tails, the claimed n^{-10} bound and the optimal O(log n) time do not follow. I am not asserting the estimate is false; I am asking the authors to make the cited result explicit and to verify that the constants are uniform in the way the proof requires.
minor comments (4)
  1. [Sec. 4.1, Lemma 4.3] The line 'Thus, L_t ⊇ L_{t-} ∪ {v}' is overstated: if v is at distance at least two from the legacy region, a flip of X_t(v) to -1 creates a new minus spacetime cluster that is not automatically absorbed into L_t. The intended inclusion D_t ⊆ L_t can still be recovered because, when v is farther than one from L_{t-}, the two Potts chains have identical neighborhoods at v and therefore cannot create a new disagreement; the proof should say this explicitly rather than asserting absorption.
  2. [Sec. 3.2, Eq. (3.8) and surrounding text] The derivation of the time derivative via the flip rates e c_{v,t} is correct in spirit, but the notation W_t^v and W_{t-} is introduced in a compressed way; it would help the reader if the definition of W_t^v were stated as a displayed equation before (3.8).
  3. [Sec. 1.1, Theorem 1] The introduction says 'O(n log n) time steps' in one place while the theorem states C log n ≤ t; this should be harmonized, since the latter is the actual claim.
  4. [Sec. 2.1, Example 2.3 and Lemma 2.4] The condition βp = 2β + C log(q-1)/d appears before the constant C is specified; the proof of Lemma 2.4 uses 7 log(q-1)/d, so the constant should be fixed consistently in the statement of Example 2.3.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the rigid-dynamics recursion is self-contained and the external inputs are independent equilibrium/spectral facts, not the dynamical conclusion.

full rationale

The paper's central estimate, Proposition 3.1, is proved by a forward temporal induction on the probability mass function rho_t under the rigid dynamics, with an explicitly constructed threshold psi_k = k^{-2} d^{-1000-100k}; psi_k is not fitted to the data it predicts, and Proposition 3.4 is the conditional strengthening that makes the induction close. Section 4 reduces quasi-equilibration to extinction of the legacy region using only the domination Lemmas 4.5-4.6, the expansion estimates of Lemmas 4.9-4.10, and the rigid-dynamics tail bounds of Section 3. The one imported equilibrium input is Eq. (4.6), the concentration of the conditional phase measure pi_1 on {m >= 1 - gamma_0}, cited to [13] for Ising and to [2,12] for Potts. That input is a statement about the stationary measure, structurally distinct from the dynamical coupling claim, and it is not derived from the paper's own conclusions; the Potts citation [2] is co-authored by one of the present authors but is used together with the independent [12], so it is not a load-bearing self-citation. No prediction in Theorems 1-2 is defined in terms of the quantity it predicts, and no fitted parameter is renamed as a prediction; the bias threshold epsilon_0 comes from a separate magnetization drift argument (Lemma 5.1) using Friedman's theorem and expansion estimates. The derivation is therefore self-contained given standard external facts, with no circular step exhibited.

Assumptions & free parameters 3 free parameters · 4 assumptions · 3 invented entities

The central claim rests on no data fitting. The hand-chosen constants C0, γ0, and tail exponents appear only in sufficient conditions and can be varied. The rigid dynamics and spacetime cluster objects are new mathematical constructions with no independent empirical handle.

free parameters (3)
  • C0
    Universal constant in the low-temperature thresholds β0 = C0 log(d)/d and βp = 2β + C0 log(q)/d; chosen sufficiently large to make drift and tail estimates work.
  • γ0 = 10^{-5}
    Small universal constant in Theorem 4.1 governing the allowed deviation from all-plus initialization; chosen to satisfy the expansion Lemma 4.10.
  • ψ tail exponents 1000 and 100
    Constants in the exponential tail function ψ_k = k^{-2}d^{-1000-100k}; authors note they can be replaced by any large constants, so they are proof machinery rather than fitted values.
assumptions (4)
  • standard math Friedman's second eigenvalue theorem for random d-regular graphs
    Used in Lemma 4.9 to bound the number of vertices with high degree into a small set; cited to [17].
  • domain assumption Concentration of the constrained Ising/Potts measure: π1(m(σ) ≥ 1-γ0) ≥ 1-e^{-cn} for βp > C log q/d
    Imported from [13] (Ising) and [2,12] (Potts) at Eq. (4.6); used to show the all-phase chain is close to π1.
  • domain assumption Random d-regular graph is 1-locally-treelike with probability 1-o(1)
    Fact 2.17 and classical configuration-model estimates; used throughout for trifurcation bounds and short-cycle control.
  • domain assumption Exponential bottleneck between phases at low temperature
    Used to define the metastable measure and to justify the exponential-time upper bound; cited to [13] and [12].
invented entities (3)
  • Minus spacetime clusters and minus regions
    purpose: Combinatorial objects tracking negative information along the spacetime history of the all-plus chain; their exponential tails are the core estimate.
    Novel mathematical constructions, not physical entities; they exist only inside the proof.
  • Rigid dynamics with trifurcation points
    purpose: Non-Markovian dominating process that suppresses flips of trifurcation points, enabling negative drift on minus region sizes.
    Novel proof device; has no direct falsifiable predictions outside the paper.
  • Legacy region L_t
    purpose: Union of minus spacetime clusters incident to initial minuses; its extinction implies coupling of biased chain to the all-plus chain.
    Definition 4.2; internal to the proof.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Rapid phase ordering for Ising and Potts dynamics on random regular graphs." pith.science (2026). https://pith.science/paper/CF2NFCO2

@misc{pith2026250515783,
  author       = {Pith},
  title        = {Pith review of: Rapid phase ordering for Ising and Potts dynamics on random regular graphs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CF2NFCO2}},
  note         = {Machine review of arXiv:2505.15783}
}
abstract

We consider the Ising, and more generally, $q$-state Potts Glauber dynamics on random $d$-regular graphs on $n$ vertices at low temperatures $\beta \gtrsim \frac{\log d}{d}$. The mixing time is exponential in $n$ due to a bottleneck between $q$ dominant phases consisting of configurations in which the majority of vertices are in the same state. We prove that for any $d\ge 7$, from biased initializations with $\epsilon_d n$ more vertices in state-$1$ than in other states, the Glauber dynamics quasi-equilibrates to the stationary distribution conditioned on having plurality in state-$1$ in optimal $O(\log n)$ time. Moreover, the requisite initial bias $\epsilon_d$ can be taken to zero as $d \to \infty$. Even for the $q=2$ Ising case, where the states are naturally identified with $\pm 1$, proving such a result requires a new approach in order to control negative information spread in spacetime despite the model being in low temperature and exhibiting strong local correlations. For this purpose, we introduce a coupled non-Markovian rigid dynamics for which a delicate temporal recursion on probability mass functions of minus spacetime cluster sizes establishes their subcriticality.

Figures

Figures reproduced from arXiv: 2505.15783 by the authors.

Figure 1
Figure 1. Spacetime diagrams for coupled Ising dynamics initialized from all-plus (left) and biased initializa [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Snapshots of an Ising dynamics spacetime diagram initialized from biased initialization [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Characterizing the limiting critical Potts measures on locally regular-tree-like expander graphs

    math.PR 2025-05 conditional novelty 8.0 of 10

    At the critical line, local weak limits of Potts and random cluster measures on locally tree-like expander graphs are exactly mixtures of the free and wired tree Gibbs measures, and any mixture weight is realizable.

  2. Better Models and Algorithms for Learning Ising Models from Dynamics

    cs.LG 2025-07 conditional novelty 7.0 of 10

    First algorithms recover Ising structure and parameters from flip-only dynamics trajectories, in time poly(d)n^2 log n for structure and O~(2^d n) for parameters.

Reference graph

Works this paper leans on

33 extracted references · 31 canonical work pages · cited by 2 Pith papers

  1. [13]

    Ising models on locally tree-like graphs

    Amir Dembo and Andrea Montanari. Ising models on locally tree-like graphs. Ann. Appl. Probab., 20(2):565–592, 2010

    work page 2010

  2. [1]

    Spectral independence in high-dimensional expanders and applications to the hardcore model

    Nima Anari, Kuikui Liu, and Shayan Oveis Gharan. Spectral independence in high-dimensional expanders and applications to the hardcore model. In 2020 IEEE 61st Annual Symposium on Foundations of Computer Science (FOCS) , pages 1319–1330, 2020

    work page 2020

  3. [2]

    Potts and random cluster measures on locally regular-tree-like graphs, 2023

    Anirban Basak, Amir Dembo, and Allan Sly. Potts and random cluster measures on locally regular-tree-like graphs, 2023

    work page 2023

  4. [3]

    Kawasaki dynamics beyond the uniqueness threshold

    Roland Bauerschmidt, Thierry Bodineau, and Benoit Dagallier. Kawasaki dynamics beyond the uniqueness threshold. Probability Theory and Related Fields , 2024

    work page 2024

  5. [4]

    Stochastic dynamics and the Polchinski equation: An introduction

    Roland Bauerschmidt, Thierry Bodineau, and Benoit Dagallier. Stochastic dynamics and the Polchinski equation: An introduction. Probability Surveys, 21(none):200 – 290, 2024

    work page 2024

  6. [5]

    Log-Sobolev inequality for near critical Ising models

    Roland Bauerschmidt and Benoit Dagallier. Log-Sobolev inequality for near critical Ising models. Communications on Pure and Applied Mathematics , 77(4):2568–2576, 2024

    work page 2024

  7. [6]

    Convergence, unanimity and disagreement in majority dynamics on unimodular graphs and random graphs

    Itai Benjamini, Siu-On Chan, Ryan O’Donnell, Omer Tamuz, and Li-Yang Tan. Convergence, unanimity and disagreement in majority dynamics on unimodular graphs and random graphs. Stochastic Processes and their Applications , 126(9):2719–2733, 2016. 39

    work page 2016

  8. [7]

    Mean-field Potts and random-cluster dynamics from high-entropy initializations, 2024

    Antonio Blanca, Reza Gheissari, and Xusheng Zhang. Mean-field Potts and random-cluster dynamics from high-entropy initializations, 2024. Extended abstract appeared in SODA 2025

    work page 2024

Show all 33 references
  1. [8]

    Theory of phase-ordering kinetics

    Alan J Bray. Theory of phase-ordering kinetics. Advances in Physics , 43(3):357–459, 1994

    1994

  2. [9]

    Phase ordering after a deep quench: The stochastic Ising and hard core gas models on a tree

    Pietro Caputo and Fabio Martinelli. Phase ordering after a deep quench: The stochastic Ising and hard core gas models on a tree. Probability Theory and Related Fields, 136(1):37–80, 2006

    2006

  3. [10]

    Localization Schemes: A Framework for Proving Mixing Bounds for Markov Chains (extended abstract)

    Yuansi Chen and Ronen Eldan. Localization Schemes: A Framework for Proving Mixing Bounds for Markov Chains (extended abstract) . In 2022 IEEE 63rd Annual Symposium on Foundations of Computer Science (FOCS), pages 110–122, Los Alamitos, CA, USA, November

    2022

  4. [11]

    Goldberg, Will Perkins, James Stewart, and Eric Vigoda

    Zongchen Chen, Andreas Galanis, Leslie A. Goldberg, Will Perkins, James Stewart, and Eric Vigoda. Fast algorithms at low temperatures via markov chains. Random Structures & Algo- rithms, 58(2):294–321, 2021

    2021

  5. [12]

    Metastability of the Potts ferromagnet on random regular graphs

    Amin Coja-Oghlan, Andreas Galanis, Leslie Ann Goldberg, Jean Bernoulli Ravelomanana, Daniel ˇStefankoviˇ c, and Eric Vigoda. Metastability of the Potts ferromagnet on random regular graphs. Communications in Mathematical Physics , 401(1):185–225, 2023

    2023

  6. [14]

    Friendly bisections of random graphs

    Asaf Ferber, Matthew Kwan, Bhargav Narayanan, Ashwin Sah, and Mehtaab Sawhney. Friendly bisections of random graphs. Communications of the American Mathematical Soci- ety, 2:380–416, 2022. Publisher Copyright: © 2022 by the author(s) under Creative Commons Attribution 3.0 Lice...

    2022

  7. [15]

    L. R. Fontes, R. H. Schonmann, and V. Sidoravicius. Stretched exponential fixation in stochas- tic Ising models at zero temperature. Communications in Mathematical Physics , 228(3):495– 518, 2002

    2002

  8. [16]

    Resolution of a conjecture on major- ity dynamics: Rapid stabilization in dense random graphs

    Nikolaos Fountoulakis, Mihyun Kang, and Tam´ as Makai. Resolution of a conjecture on major- ity dynamics: Rapid stabilization in dense random graphs. Random Structures & Algorithms , 57(4):1134–1156, 2020

    2020

  9. [17]

    A proof of Alon’s second eigenvalue conjecture

    Joel Friedman. A proof of Alon’s second eigenvalue conjecture. In Proceedings of the Thirty- Fifth Annual ACM Symposium on Theory of Computing , STOC ’03, page 720–724, New York, NY, USA, 2003. Association for Computing Machinery

    2003

  10. [18]

    Newman, and Daniel L

    Reza Gheissari, Charles M. Newman, and Daniel L. Stein. Zero-temperature dynamics in the dilute Curie–Weiss model. Journal of Statistical Physics , 172(4):1009–1028, 2018

    2018

  11. [19]

    Low-temperature Ising dynamics with random initializa- tions

    Reza Gheissari and Alistair Sinclair. Low-temperature Ising dynamics with random initializa- tions. Proceedings of the 54th Annual ACM SIGACT Symposium on Theory of Computing , page 1445–1458, 2022. Full version appeared in Annals of Applied Probability

    2022

  12. [20]

    Kirkpatrick, C

    S. Kirkpatrick, C. D. Gelatt, and M. P. Vecchi. Optimization by simulated annealing. Science, 220(4598):671–680, 1983. 40

    1983

  13. [21]

    Fast and Slow Mixing of the Kawasaki Dynamics on Bounded-Degree Graphs

    Aiya Kuchukova, Marcus Pappik, Will Perkins, and Corrine Yap. Fast and Slow Mixing of the Kawasaki Dynamics on Bounded-Degree Graphs. In Amit Kumar and Noga Ron-Zewi, editors, Approximation, Randomization, and Combinatorial Optimization. Algorithms and Techniques (APPROX/RANDO...

    2024

  14. [22]

    Levin and Y

    D. Levin and Y. Peres. Markov Chains and Mixing Times (2nd ed.) . American Mathematical Society, Providence, RI, 2017

    2017

  15. [23]

    Levin, Malwina J

    David A. Levin, Malwina J. Luczak, and Yuval Peres. Glauber dynamics for the mean-field Ising model: Cut-off, critical power law, and metastability. Probab. Theory Related Fields , 146(1-2):223–265, 2010

    2010

  16. [24]

    Stochastic interacting systems: contact, voter and exclusion processes , volume 324

    Thomas M Liggett. Stochastic interacting systems: contact, voter and exclusion processes , volume 324. springer science & Business Media, 2013

    2013

  17. [25]

    Cutoff phenomena for random walks on random regular graphs

    Eyal Lubetzky and Allan Sly. Cutoff phenomena for random walks on random regular graphs. Duke Mathematical Journal , 153(3):475 – 510, 2010

    2010

  18. [26]

    Information percolation and cutoff for the stochastic Ising model

    Eyal Lubetzky and Allan Sly. Information percolation and cutoff for the stochastic Ising model. Journal of the American Mathematical Society , 29(3):729–774, 2016

    2016

  19. [27]

    Lectures on Glauber dynamics for discrete spin models

    Fabio Martinelli. Lectures on Glauber dynamics for discrete spin models. In Lectures on probability theory and statistics (Saint-Flour, 1997) , volume 1717 of Lecture Notes in Math. , pages 93–191. Springer, Berlin, 1999

    1997

  20. [28]

    Zero-temperature Glauber dynamics on Zd

    Robert Morris. Zero-temperature Glauber dynamics on Zd. Probability Theory and Related Fields, 149(3):417–434, 2011

    2011

  21. [29]

    Exact thresholds for Ising–Gibbs samplers on general graphs

    Elchanan Mossel and Allan Sly. Exact thresholds for Ising–Gibbs samplers on general graphs. Ann. Probab., 41(1):294–328, 01 2013

    2013

  22. [30]

    Majority dynamics: The power of one

    Ashwin Sah and Mehtaab Sawhney. Majority dynamics: The power of one. Israel Journal of Mathematics, 2024

    2024

  23. [31]

    Reaching a Consensus on Random Networks: The Power of Few

    Linh Tran and Van Vu. Reaching a Consensus on Random Networks: The Power of Few. In Proceedings of RANDOM, 2020

    2020

  24. [32]

    Counting independent sets up to the tree threshold

    Dror Weitz. Counting independent sets up to the tree threshold. In Proceedings of the Thirty- Eighth Annual ACM Symposium on Theory of Computing , STOC ’06, page 140–149, New York, NY, USA, 2006. Association for Computing Machinery. 41

    2006

  25. [2022]

    IEEE Computer Society

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.