pith. sign in

arxiv: 2605.29608 · v1 · pith:MEV2JD6Dnew · submitted 2026-05-28 · 🧮 math.PR

Log-Sobolev Inequality for Wolff Dynamics and Application to the Condensation of Eigen Microstate in the 1D Ising Model

Kaiyuan Cui , Fuzhou Gong This is my paper

Pith reviewed 2026-06-29 05:57 UTC · model grok-4.3

classification 🧮 math.PR
keywords log-Sobolev inequalityWolff dynamics1D Ising modelMarkov chaineigen microstate condensationsubcritical regimeergodic averagessample covariance spectrum
0
0 comments X

The pith

Wolff dynamics on the 1D Ising model obeys a log-Sobolev inequality with explicit constant throughout the subcritical regime.

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

The authors first compute the exact transition probabilities of the Wolff dynamics on the one-dimensional Ising chain. With those probabilities in hand they prove a log-Sobolev inequality that holds uniformly for all temperatures below criticality and obtain explicit quantitative bounds on ergodic averages of the chain. At infinite temperature the constant recovers the classical value known for the hypercube random walk. The same bounds are then applied to the sample covariance matrix generated by the dynamics, establishing that its spectrum exhibits the condensation of eigen microstates previously seen only in simulations. A reader cares because the result supplies the first rigorous mixing-time control for this widely used non-local sampler in a regime where critical slowing down is absent.

Core claim

For the one-dimensional Ising model the transition probabilities of the Wolff dynamics admit a closed-form expression. Substituting these expressions yields a log-Sobolev inequality with an explicit constant that remains valid for every inverse temperature strictly less than the critical value. The resulting control on ergodic averages is used to prove that the spectrum of the sample covariance matrix produced by the dynamics agrees with the condensation behavior observed numerically by Chen et al.

What carries the argument

The closed-form transition probabilities of the Wolff dynamics on the 1D Ising chain, which permit direct verification of the log-Sobolev inequality.

If this is right

  • The Wolff chain mixes at a rate controlled by the explicit constant for every subcritical temperature.
  • At infinite temperature the constant coincides with the known log-Sobolev constant of the hypercube random walk.
  • Ergodic averages of observables under the Wolff dynamics satisfy explicit quantitative bounds.
  • The spectrum of the sample covariance matrix generated by the dynamics exhibits condensation of eigen microstates, matching prior simulations.

Where Pith is reading between the lines

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

  • The same explicit-probability technique may allow log-Sobolev constants to be derived for other one-dimensional spin systems with non-local updates.
  • The oscillation between the two fully aligned states at criticality suggests that related non-local chains may exhibit periodic limiting behavior.
  • The explicit constant could be inserted into existing variance bounds for Monte Carlo estimators that go beyond the covariance spectrum studied here.

Load-bearing premise

The log-Sobolev proof depends on having explicit closed-form transition probabilities, which exist only because the model is one-dimensional.

What would settle it

Compute the Dirichlet form and variance ratio for the Wolff chain at a fixed subcritical temperature on a chain of length 20 and check whether the ratio equals the claimed explicit constant.

read the original abstract

The Wolff dynamics is a non-local Markov chain widely used for simulating the Ising model due to its effectiveness in reducing critical slowing down compared to the Glauber dynamics. Despite extensive algorithmic and numerical studies, a rigorous probabilistic understanding remains limited. In this paper, we take a first step toward addressing this gap. For the one-dimensional (1D) Ising model, we first derive the transition probabilities of the Wolff dynamics and show that, at the critical point, it converges to the two fully aligned configurations and subsequently oscillates between them. This behavior is absent in the Glauber dynamics. Second, we establish a log-Sobolev inequality with an explicit constant for the Wolff dynamics in the entire subcritical regime and derive quantitative bounds on its ergodic averages. As a by-product, at infinite temperature, the obtained constant coincides with the classical log-Sobolev constant of the random walk on the hypercube. Finally, we apply these results to analyze the spectrum of the sample covariance matrix generated by the Wolff dynamics, which was used by Chen et al. to study condensation of eigen microstate. We prove that the spectral behavior agrees with their simulations in the 1D Ising model, thereby providing theoretical support for their findings.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript derives explicit closed-form transition probabilities for the Wolff dynamics on the one-dimensional Ising chain. It shows that at criticality the chain converges to the pair of fully aligned configurations and oscillates between them. It then proves a log-Sobolev inequality with an explicit constant throughout the subcritical regime, obtains quantitative bounds on ergodic averages, and notes that the constant recovers the classical hypercube random-walk value at infinite temperature. As an application, the authors analyze the spectrum of the sample covariance matrix generated by the dynamics and prove agreement with the condensation-of-eigen-microstate observations reported in simulations by Chen et al.

Significance. If the explicit transition probabilities are correct, the work supplies the first rigorous probabilistic analysis of Wolff dynamics in 1D, including an explicit LSI constant and ergodic bounds that are uniform in the subcritical regime. The recovery of the known hypercube constant at infinite temperature provides an independent consistency check. The spectral application furnishes theoretical support for the numerical findings of Chen et al. on eigen-microstate condensation. The explicit, parameter-free character of the derivations is a clear strength.

major comments (2)
  1. [§3 (transition-probability derivation)] The entire LSI proof (and therefore the ergodic bounds and the spectral application) rests on the correctness of the closed-form transition probabilities derived from the linear cluster structure. Any algebraic or combinatorial error in those expressions would invalidate the subsequent claims. The manuscript should supply an independent verification—e.g., direct enumeration for small chain lengths or comparison against the known Glauber transition matrix in a suitable limit—to confirm the formulas before the LSI step.
  2. [§3.2 and §4] The statement that the dynamics “converges to the two fully aligned configurations and subsequently oscillates between them” at criticality is used to motivate the subcritical analysis. The precise sense of convergence and the period-2 oscillation must be stated rigorously (e.g., in total variation or in the LSI norm) and shown to be compatible with the LSI constant derived for β < β_c.
minor comments (2)
  1. [§2] Notation for the Wolff cluster flip probability and the resulting transition kernel should be introduced once and used consistently; several passages reuse the same symbol for the single-spin and cluster-level kernels.
  2. [§5] The comparison with Chen et al. would benefit from an explicit statement of which spectral quantities (eigenvalue gap, eigenvector localization, etc.) are being matched and the precise regime of β in which the agreement is claimed.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and valuable suggestions, which will help improve the clarity and rigor of the manuscript. We address the major comments point by point below.

read point-by-point responses

  1. Referee: [§3 (transition-probability derivation)] The entire LSI proof (and therefore the ergodic bounds and the spectral application) rests on the correctness of the closed-form transition probabilities derived from the linear cluster structure. Any algebraic or combinatorial error in those expressions would invalidate the subsequent claims. The manuscript should supply an independent verification—e.g., direct enumeration for small chain lengths or comparison against the known Glauber transition matrix in a suitable limit—to confirm the formulas before the LSI step.

    Authors: We agree that an independent check strengthens the foundation. In the revised version we will add, in a new subsection of §3, explicit transition matrices computed by direct enumeration for n=2 and n=3, together with a side-by-side comparison against the closed-form expressions. We will also record the β→0 limit, which recovers the hypercube random-walk transitions and is already consistent with the recovery of the classical LSI constant stated in the paper. revision: yes

  2. Referee: [§3.2 and §4] The statement that the dynamics “converges to the two fully aligned configurations and subsequently oscillates between them” at criticality is used to motivate the subcritical analysis. The precise sense of convergence and the period-2 oscillation must be stated rigorously (e.g., in total variation or in the LSI norm) and shown to be compatible with the LSI constant derived for β < β_c.

    Authors: We accept that the critical-case description needs a precise formulation. In the revision we will insert, at the end of §3.2, a rigorous statement: at β=β_c the total-variation distance from the law of the chain to the two-point set {all-up, all-down} tends to zero, while the chain alternates deterministically between the two configurations on even and odd steps. Because the log-Sobolev inequality and all quantitative bounds are proved only for β<β_c, we will add an explicit remark that the critical regime is treated separately and does not invoke the subcritical LSI constant; the subcritical results remain valid uniformly as β approaches β_c from below. revision: yes

Circularity Check

0 steps flagged

No circularity: LSI derived from explicit transition probabilities, independent of self-citation or fitting

full rationale

The paper computes closed-form transition probabilities for Wolff dynamics on the 1D Ising chain, then uses these to prove an explicit LSI in the subcritical regime and bound ergodic averages. The spectral application confirms agreement with external Chen et al. simulations. No step reduces by construction to its inputs, no fitted parameters are relabeled as predictions, and no load-bearing self-citation or uniqueness theorem is invoked. The derivation chain is self-contained.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work rests on standard Markov-chain axioms and the explicit solvability of the 1D Ising model; no free parameters, new entities, or ad-hoc assumptions are introduced in the abstract.

axioms (1)
  • standard math Standard properties of reversible Markov chains and the definition of the log-Sobolev inequality apply to the Wolff transition kernel.
    Invoked to obtain the explicit constant from the computed transition probabilities.

pith-pipeline@v0.9.1-grok · 5753 in / 1183 out tokens · 24472 ms · 2026-06-29T05:57:45.128923+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

46 extracted references · 41 canonical work pages

  1. [1]

    and CAO, S

    ADHIKARI, A. and CAO, S. (2025) Correlation decay for finite lattice gauge theories at weak coupling. Ann. Probab.53140–174. MR4852890 https://doi.org/10.1214/24-AOP1702

    work page doi:10.1214/24-aop1702 2025

  2. [2]

    and LEDOUX, M

    BAKRY, D., GENTIL, I. and LEDOUX, M. (2014).Analysis and Geometry of Markov Diffusion Op- erators. Grundlehren der mathematischen Wissenschaften [Fundamental Principles of Mathematical Sciences]348. Springer, Cham. MR3155209 https://doi.org/10.1007/978-3-319-00227-9

    work page doi:10.1007/978-3-319-00227-9 2014

  3. [3]

    and DIACONIS, P

    BASSETTI, F. and DIACONIS, P. (2006). Examples comparing importance sampling and the Metropolis algorithm.Illinois J. Math.5067–91. MR2247824 https://doi.org/10.1215/ijm/1258059470

    work page doi:10.1215/ijm/1258059470 2006

  4. [4]

    and BODINEAU, T

    BAUERSCHMIDT, R. and BODINEAU, T. (2021). Log-Sobolev inequality for the continuum sine- Gordon model.Comm. Pure Appl. Math.742064–2113. MR4303014 https://doi.org/10.1002/cpa.21926

    work page doi:10.1002/cpa.21926 2021

  5. [5]

    and DAGALLIER, B

    BAUERSCHMIDT, R. and DAGALLIER, B. (2024). Log-Sobolev inequality for near critical Ising mod- els.Comm. Pure Appl. Math.772568–2576. MR4705299 https://doi.org/10.1002/cpa.22172

    work page doi:10.1002/cpa.22172 2024

  6. [6]

    and DAGALLIER, B

    BAUERSCHMIDT, R. and DAGALLIER, B. (2024). Log-Sobolev inequality for theϕ 4 2 andϕ 4 3 measures. Comm. Pure Appl. Math.772579–2612. MR4720217 https://doi.org/10.1002/cpa.22173

    work page doi:10.1002/cpa.22173 2024

  7. [7]

    and VIGODA, E

    BLANCA, A., CAPUTO, P., PARISI, D., SINCLAIR, A. and VIGODA, E. (2022). Entropy de- cay in the Swendsen–Wang dynamics onZ d.Ann. Appl. Probab.321018–1057. MR4414700 https://doi.org/10.1214/21-AAP1702 29

    work page doi:10.1214/21-aap1702 2022

  8. [8]

    and HELFFER, B

    BODINEAU, T. and HELFFER, B. (1999). The log-Sobolev inequality for unbounded spin systems.J. Funct. Anal.166168–178. MR1704666 https://doi.org/10.1006/jfan.1999.3419

    work page doi:10.1006/jfan.1999.3419 1999

  9. [9]

    and CHEN, X

    CHEN, X., YING, N., CHEN, D., ZHANG, Y., LU, B., FAN, J. and CHEN, X. (2021). Eigen microstates and their evolution of global ozone at different geopotential heights.Chaos31071102. MR4285719 https://doi.org/10.1063/5.0058599

    work page doi:10.1063/5.0058599 2021

  10. [10]

    and ELDAN, R

    CHEN, Y. and ELDAN, R. (2025). Localization schemes: A framework for proving mixing bounds for Markov chains.Duke Math. J.1741431–1510. MR4916109 https://doi.org/10.1215/00127094-2024- 0063

    work page doi:10.1215/00127094-2024- 2025

  11. [11]

    and SALOFF-COSTE, L

    DIACONIS, P. and SALOFF-COSTE, L. (1996). Logarithmic Sobolev inequalities for finite Markov chains.Ann. Appl. Probab.6695–750. MR1410112 https://doi.org/10.1214/aoap/1034968224

    work page doi:10.1214/aoap/1034968224 1996

  12. [12]

    and SUN, R

    DING, J., SONG, J. and SUN, R. (2023). A new correlation inequality for Ising models with external fields.Probab. Theory Related Fields186477–492. MR4586225 https://doi.org/10.1007/s00440-022- 01132-1

    work page doi:10.1007/s00440-022- 2023

  13. [13]

    and RAOUFI, A

    DUMINIL-COPIN, H., GOSWAMI, S. and RAOUFI, A. (2020). Exponential decay of truncated corre- lations for the Ising model in any dimension for all but the critical temperature.Commun. Math. Phys. 374891–921. MR4072233 https://doi.org/10.1007/s00220-019-03633-y

    work page doi:10.1007/s00220-019-03633-y 2020

  14. [14]

    and ZEITOUNI, O

    ELDAN, R., KOEHLER, F. and ZEITOUNI, O. (2022). A spectral condition for spectral gap: fast mixing in high-temperature Ising models.Probab. Theory Related Fields1821035–1051. MR4408509 https://doi.org/10.1007/s00440-021-01085-x

    work page doi:10.1007/s00440-021-01085-x 2022

  15. [15]

    and WU, L

    GONG, F. and WU, L. (2006). Spectral gap of positive operators and applications.J. Math. Pures Appl.85151–191. MR2199011 https://doi.org/10.1016/j.matpur.2004.11.004

    work page doi:10.1016/j.matpur.2004.11.004 2006

  16. [16]

    GROSS, L. (1975). Logarithmic Sobolev inequalities.Amer. J. Math.971061–1083. MR0420249 https://doi.org/10.2307/2373688

    work page doi:10.2307/2373688 1975

  17. [17]

    HELFFER, B. (1998). Remarks on decay of correlations and Witten Laplacians, Brascamp- Lieb inequalities and semiclassical limit.J. Funct. Anal.155571–586. MR1624506 https://doi.org/10.1006/jfan.1997.3239

    work page doi:10.1006/jfan.1997.3239 1998

  18. [18]

    HELFFER, B. (1999). Remarks on decay of correlations and Witten Laplacians. III. Application to logarithmic Sobolev inequalities.Ann. Inst. H. Poincar ´e Probab. Statist.35483–508. MR1702239 https://doi.org/10.1016/S0246-0203(99)00103-X

    work page doi:10.1016/s0246-0203(99)00103-x 1999

  19. [19]

    and CHEN, X (2019)

    HU, G., LIU, T., LIU, M., CHEN, W. and CHEN, X (2019). Condensation of eigen mi- crostate in statistical ensemble and phase transition.Sci. China Phys. Mech. Astron.62990511. https://doi.org/10.1007/s11433-018-9353-x

    work page doi:10.1007/s11433-018-9353-x 2019

  20. [20]

    and CHEN, X

    HU, G., LIU, M. and CHEN, X. (2023). Quantum phase transition and eigen mi- crostate condensation in the quantum Rabi model.Phys. A630129210. MR4647805 https://doi.org/10.1016/j.physa.2023.129210

    work page doi:10.1016/j.physa.2023.129210 2023

  21. [21]

    ISING, E. (1925). Beitrag zur theorie des ferromagnetismus.Phys. Zeitschr.31253–258. https://doi.org/10.1007/BF02980577 30

    work page doi:10.1007/bf02980577 1925

  22. [22]

    (2001).Logarithmic Sobolev inequalities for unbounded spin systems revisited.In S´eminaire de Probabilit´es XXXV

    LEDOUX, M. (2001).Logarithmic Sobolev inequalities for unbounded spin systems revisited.In S´eminaire de Probabilit´es XXXV . Lecture Notes in Math.1755167–194. Springer, Berlin. MR1837286 https://doi.org/10.1007/978-3-540-44671-2 13

    work page doi:10.1007/978-3-540-44671-2 2001

  23. [23]

    LENZ, W. (1920). Beitrag zum Verst ¨andnis der magnetischen Erscheinungen in festen K¨orpern.Phys. Zeitschr.21613–615

    1920

  24. [24]

    LEVIN, D. A. and PERES, Y. (2017).Markov chains and mixing times.Amer. Math. Soc., Providence, RI. MR3726904

    2017

  25. [25]

    K., DONG, J

    LIU, T., HU, G. K., DONG, J. Q., FAN, J. F., LIU, M. X. and CHEN, X. S. (2022). Renormal- ization group theory of eigen microstates.Chin. Phys. Lett.39080503. https://doi.org/10.1088/0256- 307X/39/8/080503

    work page doi:10.1088/0256- 2022

  26. [26]

    (2006).Introduction to Cluster Monte Carlo Algorithms

    LUIJTEN, E. (2006).Introduction to Cluster Monte Carlo Algorithms. In Computer Simulations in Condensed Matter Systems: From Materials to Chemical Biology V olume 1.70313–38. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-35273-2

    work page doi:10.1007/3-540-35273-2 2006

  27. [27]

    MARTINELLI, F. (1997). Lectures on Glauber dynamics for discrete spin models. In Lectures on prob- ability theory and statistics (Saint-Flour, 1997). Lecture Notes in Math.171793–191. Springer, Berlin. MR1746301 https://doi.org/10.1007/978-3-540-48115-7 2

    work page doi:10.1007/978-3-540-48115-7 1997

  28. [28]

    MENZ, G. (2014). A Brascamp-Lieb type covariance estimate.Electron. J. Probab.191–15. MR3256878 https://doi.org/10.1214/EJP.v19-2997

    work page doi:10.1214/ejp.v19-2997 2014

  29. [29]

    MENZ, G. (2014). The approach of Otto-Reznikoff revisited.Electron. J. Probab.191–27. MR3275859 https://doi.org/10.1214/EJP.v19-3418

    work page doi:10.1214/ejp.v19-3418 2014

  30. [30]

    and ERCOLESSI, E

    MORANDI, G., NAPOLI, F. and ERCOLESSI, E. (2004).Statistical Mechanics: An Intermediate Course.World Scientific Publishing Co. Pte. Ltd, Singapore

    2004

  31. [31]

    and SLY, A

    NAM, D. and SLY, A. (2019). Cutoff for the Swendsen–Wang dynamics on the lattice.Ann. Probab. 473705–3761. MR4038041 https://doi.org/10.1214/19-AOP1344

    work page doi:10.1214/19-aop1344 2019

  32. [32]

    and VELENIK, Y

    OTT, S. and VELENIK, Y. (2023).Asymptotics of correlations in the Ising model: a brief survey. Topics in statistical mechanics, 157–184, Panor. Synth`eses, 59, Soc. Math. France, Paris. MR4619313

    2023

  33. [33]

    and REZNIKOFF, M

    OTTO, F. and REZNIKOFF, M. G. (2007). A new criterion for the logarithmic Sobolev inequality and two applications.J. Funct. Anal.243121–157. MR2291434 https://doi.org/10.1016/j.jfa.2006.10.002

    work page doi:10.1016/j.jfa.2006.10.002 2007

  34. [34]

    (2007).An initiation to logarithmic Sobolev inequalities.SMF/AMS Texts and Mono- graphs14

    ROYER, G. (2007).An initiation to logarithmic Sobolev inequalities.SMF/AMS Texts and Mono- graphs14. Amer. Math. Soc., Providence, RI. MR2352327

    2007

  35. [35]

    RUDOLF, D. (2010). Error bounds for computing the expectation by Markov chain Monte Carlo.Monte Carlo Methods Appl.16323–342. MR2747819 https://doi.org/10.1515/mcma.2010.012

    work page doi:10.1515/mcma.2010.012 2010

  36. [36]

    and ZHU, X

    SHEN, H., ZHU, R. and ZHU, X. (2023) A stochastic analysis approach to lattice Yang–Mills at strong coupling.Comm. Math. Phys.400805–851. MR4589716 https://doi.org/10.1007/s00220-022-04609-1

    work page doi:10.1007/s00220-022-04609-1 2023

  37. [37]

    and ZHU, X

    SHEN, H., ZHU, R. and ZHU, X. (2026) Langevin dynamics of lattice Yang-Mills-Higgs and applica- tions.Comm. Math. Phys.407MR5016805 https://doi.org/10.1007/s00220-025-05528-7 31

    work page doi:10.1007/s00220-025-05528-7 2026

  38. [38]

    SUN, Y., HU, G., ZHANG, Y., LU, B., LU, Z., FAN, J., LI, X., DENG, Q., and CHEN, X. (2021). Eigen microstates and their evolutions in complex systems.Commun. Theor. Phys.73065603. MR4269009 https://doi.org/10.1088/1572-9494/abf127

    work page doi:10.1088/1572-9494/abf127 2021

  39. [39]

    SWENDSEN, R. H. and WANG, J. S. (1987). Nonuniversal critical dynamics in Monte Carlo simula- tions.Phys. Rev. Lett.5886. https://doi.org/10.1103/PhysRevLett.58.86

    work page doi:10.1103/physrevlett.58.86 1987

  40. [40]

    and KLEIN, W

    TAMAYO, P., BROWER, R.C. and KLEIN, W. (1990). Single-cluster Monte Carlo dynamics for the Ising model.J. Statist. Phys.581083–1094. MR1049057 https://doi.org/10.1007/BF01026564

    work page doi:10.1007/bf01026564 1990

  41. [41]

    ULLRICH, M. (2013). Comparison of Swendsen-Wang and heat-bath dynamics.Random Structures Algorithms42520–535. MR3068036 https://doi.org/10.1002/rsa.20431

    work page doi:10.1002/rsa.20431 2013

  42. [42]

    ULLRICH, M. (2014). Rapid mixing of Swendsen–Wang dynamics in two dimensions.Dissertationes Math.5023–64. MR3222829 https://doi.org/10.4064/dm502-0-1

    work page doi:10.4064/dm502-0-1 2014

  43. [43]

    WEYL, H. (1912). Das asymtotische Verteilungsgesetz der Eigenwerte lineare partieller Differential- gleichungen,Math. Ann.71441–479. MR1511670 https://doi.org/10.1007/BF01456804

    work page doi:10.1007/bf01456804 1912

  44. [44]

    WOLFF, U. (1989). Collective Monte Carlo updating for spin systems.Phys. Rev. Lett.62361. https://doi.org/10.1103/PhysRevLett.62.361

    work page doi:10.1103/physrevlett.62.361 1989

  45. [45]

    YOSHIDA, N. (1999). The log-Sobolev inequality for weakly coupled lattice fields.Probab. Theory Related Fields1151–40. MR1715549 https://doi.org/10.1007/s004400050235

    work page doi:10.1007/s004400050235 1999

  46. [46]

    and CHEN, X

    ZHANG, Y., LIU, M., HU, G., LIU, T. and CHEN, X. (2024). Eigen microstates in self-organized criticality.Phys. Rev. E109044130. MR4742677 https://doi.org/10.1103/PhysRevE.109.044130 32

    work page doi:10.1103/physreve.109.044130 2024