Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

Sequential Dynamics in Ising Spin Glasses

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's central claim is that a block approximation gives the exact asymptotic law of sequential Ising spin glass dynamics, and that the same law governs the systematic scan.

desk verdict A real theorem about block-update DMFT, sold with a conjecture about the sequential scan that the paper itself does not prove. read the letter →

arxiv 2506.09877 v1 pith:S3NTLSYF submitted 2025-06-11 cond-mat.dis-nn math-phmath.MPmath.PR

Yatin Dandi , David Gamarnik , Francisco Pernice , Lenka Zdeborová This is my paper · ORCID
classification cond-mat.dis-nnmath-phmath.MPmath.PR MSC 82B4460K3582C31
keywords Sherrington-KirkpatrickmodelIsingspinglassdynamicssystematicscanGlauberdynamicalmeanfieldtheoryintegro-differenceequationsblockapproximationGaussianconditioning
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 aims to give the first exact asymptotic description of sequential, one-spin-at-a-time update dynamics on the Sherrington-Kirkpatrick model with Ising spins. It proves a precise limit theorem for a block version of the scan in which spins are updated in groups of size $\delta N$, and it derives the equations for the limiting law of spin and field histories in the regime where the number of scans is constant while $N\to\infty$ and then $\delta\to0$. The authors conjecture, and verify numerically, that the same equations describe the genuine systematic scan without blocks, including Glauber updates at any temperature. This matters because such a discrete, asynchronous analogue of the Cugliandolo-Kurchan equations for spherical spin glasses has been missing for decades.

What carries the argument

The load-bearing construction is the $\delta$-block systematic scan (Algorithm 1), which splits the spins into $1/\delta$ consecutive blocks of size $\delta N$ and updates each block in parallel while scanning blocks in a fixed order. Iterative Gaussian conditioning decomposes the Gaussian disorder $J$ into fresh independent noise plus terms lying in the span of past spin blocks, leading to a low-dimensional Effective Process (equations (56)-(57)) whose entries are approximately i.i.d. across spins at fixed $\delta$. The limit $\delta\to0$ converts the fixed-$\delta$ difference equations into the ODE/integro-difference system (4)-(11), whose unique global solution is established by Arzelà-Ascoli compactness and a Cauchy-Lipschitz uniqueness argument. The correlation function $C_{s,t}=\mathbb{E}[\sigma_s\sigma_t]$, the response function $R_{s,t}=\mathbb{E}[G_s\sigma_t]$, and the recursion for $A$ are the objects that carry the accumulated inter-spin interactions.

What would settle it

Run the systematic scan and the $\delta$-block algorithm at the same large $N$ and fixed pass count $T$, and compare the energy per spin (or two-time overlap) as $\delta$ shrinks; if the extrapolated block value differs from the scan value by more than the numerical error of the equations, the conjectured equality of limits fails.

Watch

Extended reading notes

Core claim

On its own terms, the paper claims that the macroscopic spin-field evolution of a block version of the systematic scan has an exact large-$N$ limit: for every pseudo-Lipschitz $\varphi$, the empirical block average converges, first in probability as $N\to\infty$ and then as $\delta\to0$, to $\mathbb{E}[\varphi(\sigma_x^{(T)}, h_x^{(T)})]$, where the limiting law is the unique solution of the integro-difference system (4)-(11). The law is described through correlation and response matrices $C$ and $R$ plus a recursive upper-triangular matrix $A$, so observables such as energy, magnetization and two-time overlap can be read off directly. The authors further claim, as a conjecture verified numerically in Section 7, that this $\delta\to0$ block limit equals the dynamics of the true sequential systematic scan without blocks; the proved theorem is for the block algorithm.

Load-bearing premise

The load-bearing premise is that the block dynamics with smaller and smaller blocks converges to the true sequential scan; the paper states this as a conjecture, not a theorem, so the headline claim about sequential dynamics depends on it.

Editorial extensions

If this is right

  • The limiting energy, magnetization, and two-time overlap of the block dynamics are computable from the equations, and an explicit formula is given for the energy after one Glauber pass.
  • All macroscopic observables concentrated as functions of the empirical spin-field measure; the expected values are given by averaging the block-position limit law $\nu_{x,T}$ over $x\in[0,1]$.
  • The system is numerically tractable, and Section 7's solutions match direct simulations of the systematic scan, including finite-temperature energies and zero-temperature correlations, supporting the conjecture.
  • Setting $\delta=1$ recovers the previously established equations for fully parallel dynamics, giving a unified treatment of synchronous and block-scan protocols.
  • If the conjecture is correct, the systematic scan with Glauber updates at any inverse temperature becomes the discrete, asynchronous analogue of the Cugliandolo-Kurchan equations for spherical spin glasses, filling a long-standing gap.

Reading between the lines

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

  • An editorial consequence is that a proof of the exchanged limits would elevate the sequential-scan description from numerical conjecture to theorem; until then statements about the scan inherit the conjecture.
  • The block machinery suggests a template for random-site asynchronous Glauber dynamics, where the same DMFT-style equations may hold with the scan position $x$ replaced by continuous time.
  • The apparent polynomial decay of $1-C_{t,t-1}$ at zero temperature, flagged by the authors as surprising, indicates a dynamical exponent that future work could derive from the equations.
  • Because the equations characterize all times in the linear time scale, they make algorithmic questions about local search on the hypercube empirically tractable, such as what energy a given scan order can reach.
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 discrete sequential spin updates on the Sherrington-Kirkpatrick model with Ising spins. The authors introduce a δ-block variant of the systematic scan (Algorithm 1), in which blocks of δN spins are updated in parallel while blocks are scanned sequentially, and prove a limit theorem: for pseudo-Lipschitz test functions, the empirical measure of spin-field histories along the block dynamics converges in the double limit N→∞ then δ→0 to the law described by the integro-difference system (4)-(11). The proof uses iterative Gaussian conditioning, concentration of low-dimensional order parameters, and an Arzelà-Ascoli argument for the δ→0 convergence of the block equations. The paper also presents numerical solutions of the limiting equations and compares them with simulations of the ordinary one-spin-at-a-time systematic scan. The authors state explicitly in Section 2 and Section 1.3 that the identification of the δ→0 block dynamics with the systematic scan is a conjecture deferred to future work, verified only experimentally in Section 7.

Significance. The block-dynamics theorem is a substantial technical contribution: it gives a rigorous, parameter-free DMFT-type description of a coarse-grained sequential update process on the SK model, with explicit limits for the energy, magnetization, and correlations, and it recovers the parallel (δ=1) case studied earlier. The derivation is detailed and structurally sound, and the limiting equations are numerically tractable. However, the significance as advertised in the title and abstract—an exact asymptotic characterization of sequential (systematic scan) dynamics resolving a long-standing gap—is not supported by the proved results, because Theorem 1 concerns the δ-block process and the δ→0 identification with the true sequential scan is only conjectured. The paper's value is therefore best assessed as a rigorous analysis of block-sequential dynamics plus a well-posed conjecture for the sequential limit, rather than as a proof of the sequential claim.

major comments (3)
  1. [Section 2, Theorem 1; Section 1.3] The abstract and title claim an exact characterization of sequential dynamics implemented via systematic scan, but the theorem proved concerns Algorithm 1, the δ-block systematic scan, in the double limit lim_{δ→0} lim_{N→∞}. Section 2 states: 'We conjecture that this agrees with the limiting dynamics of the systematic scan itself... we will address this conjecture in upcoming work, but for now we verify it experimentally in Section 7.' Section 1.3 similarly defers the identification. This gap is load-bearing for the advertised sequential claim: Theorem 1 and Corollary 1 do not prove statements about the one-spin-at-a-time scan. The manuscript should either prove the δ→0 identification or, at minimum, reframe the main theorem and the abstract around the block process and present the sequential claim as a conjecture.
  2. [Appendix A.1, Proposition 5 (inequalities (99)-(100))] The proof of Proposition 5 contains an invalid estimate. After deriving |f^{t,j}|_2 ≤ δ Σ_{k<j} B''_{t-1} h_2(δk) + B''_{t-1}, the text claims this is at most (1/2)∫_0^{δj} B''_{t-1} h_2(y)dy + B''_{t-1}. But h_2 defined by the ODE (101) is increasing, so for increasing positive h_2 the left Riemann sum δ Σ_{k<j} h_2(δk) is at least (not at most) half the integral; indeed it is typically close to the full integral. Since Proposition 5 provides the uniform δ-independence of f, Σ, C, A needed for Theorem 4 and hence Theorem 1, this gap is load-bearing. The proof is likely repairable (e.g., setting h_1'(x)=B''_{t-1}h_2(x) and using the standard left-sum bound), but the argument as written does not establish the proposition.
  3. [Theorem 1 statement] The block indexing in Theorem 1 is inconsistent with the definition in Section 2. Algorithm 1 and the text define B(j) = {i∈[N]: (j-1)δN < i ≤ jδN}, but Theorem 1 averages over i=jδN,...,(j+1)δN-1, which is block j+1 under that convention, and for x=1 (where j=⌊1/δ⌋=1/δ) the range extends beyond N. The theorem should be restated with a consistent 0- or 1-indexed block definition, for example averaging over {i: jδN < i ≤ (j+1)δN} with j=⌊x/δ⌋ and x∈[0,1), or by defining B(j) accordingly.
minor comments (4)
  1. [Section 1.1, paragraph on parallel vs sequential dynamics] The sentence 'in the latter case the energy is non-decreasing in time' is not correct for the general Glauber dynamics at finite temperature considered in the paper; energy monotonicity holds for zero-temperature energy-increasing updates but not for stochastic Glauber updates at positive temperature.
  2. [Section 7] The experimental comparison uses N=1500 and no error bars or finite-size scaling. Since Section 7 is the only evidence for the conjectured δ→0 identification with the systematic scan, reporting multiple system sizes and standard errors would materially strengthen the claim.
  3. [Theorem 1 and Definition 1] Theorem 1 says 'T>0' and 'pseudo-Lipschitz of finite order in the second argument'; T should be specified as a positive integer (it indexes passes), and the pseudo-Lipschitz notion should reference the order k of Definition 1 explicitly.
  4. [Throughout, notation] The convention for A_{0:-1,0} and C_{0,0}=1 is used in equation (10)-(11) and below; it is stated, but the zero-indexing conventions for matrices are intricate and would benefit from a summary table or a figure, as errors are easy for readers to make.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: Theorem 1 is proved for δ-block dynamics; the systematic-scan identification is an explicitly stated conjecture, not a fitted or definitional reduction.

full rationale

The paper's derivation chain is self-contained: the effective process (4)-(11) is not defined in terms of the predicted observables, but rather constructed from a Gaussian conditioning decomposition of the GOE disorder, with order parameters fixed by DMFT-style self-consistency. Theorem 2 establishes convergence of the δ-block dynamics to this effective process as N→∞, and Theorem 4 establishes the δ→0 limit of the block equations. No parameter is fitted to simulation data; the numerical comparisons in Section 7 are tests, not inputs. The only substantial gap is that the advertised identification with the systematic scan is explicitly left as a conjecture: the paper states, "We conjecture that this agrees with the limiting dynamics of the systematic scan itself (without block updates)... we will address this conjecture in an upcoming work, but for now we verify it experimentally in Section 7." This is an honest limitation and an unproven limit, not a circular reduction, because the double limit for Algorithm 1 is proved independently of that conjecture. The self-citation to [EBKZ24] is used only as a comparison for the δ=1 parallel case and is not load-bearing for the main theorem. Hence no circular step is present.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The central derivation rests on the GOE Gaussian structure, the non-degeneracy of the update rule, the fixed-T regime, and classical compactness theorems. No free parameters are fitted to data, and no new physical entities are postulated.

assumptions (4)
  • domain assumption Update rule c is non-degenerate: 0 < Pr[c(G)=1] < 1 for standard Gaussian G, and the boundary of c^{-1}(1) has Lebesgue measure zero
    Assumed in Section 2 to avoid ties and ensure closeness of spins under coupled fields (Lemma 8); excludes some pathological update rules.
  • domain assumption Time horizon T is fixed as N→∞ (linear-time regime)
    The theorem and equations are for T constant; no result is claimed for T growing with N.
  • standard math GOE covariance structure of the disorder matrix J
    Definition of the SK model in Section 1; the proof relies on Gaussian conditioning and Stein's lemma for Gaussian vectors.
  • standard math Arzela-Ascoli theorem and Cauchy-Lipschitz theorem
    Used in Section 5 to establish convergence and uniqueness of the δ→0 limit equations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Sequential Dynamics in Ising Spin Glasses." pith.science (2026). https://pith.science/paper/S3NTLSYF

@misc{pith2026250609877,
  author       = {Pith},
  title        = {Pith review of: Sequential Dynamics in Ising Spin Glasses},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/S3NTLSYF}},
  note         = {Machine review of arXiv:2506.09877}
}
read the original abstract

We present the first exact asymptotic characterization of sequential dynamics for a broad class of local update algorithms on the Sherrington-Kirkpatrick (SK) model with Ising spins. Focusing on dynamics implemented via systematic scan -- encompassing Glauber updates at any temperature -- we analyze the regime where the number of spin updates scales linearly with system size. Our main result provides a description of the spin-field trajectories as the unique solution to a system of integro-difference equations derived via Dynamical Mean Field Theory (DMFT) applied to a novel block approximation. This framework captures the time evolution of macroscopic observables such as energy and overlap, and is numerically tractable. Our equations serve as a discrete-spin sequential-update analogue of the celebrated Cugliandolo-Kurchan equations for spherical spin glasses, resolving a long-standing gap in the theory of Ising spin glass dynamics. Beyond their intrinsic theoretical interest, our results establish a foundation for analyzing a wide variety of asynchronous dynamics on the hypercube and offer new avenues for studying algorithmic limitations of local heuristics in disordered systems.

Figures

Figures reproduced from arXiv: 2506.09877 by the authors.

Figure 1
Figure 1. Energy as a function of t ′ . The faint colored curves are obtained by numerically solv￾ing our equations, while the thinner, darker curves overlaid on top are obtained by averaging 50 simulation samples with N = 1500. Dotted vertical lines mark the transitions between passes. At higher temperatures the simulation curves are noisier; at lower temperatures the two curves almost perfectly overlap. 44 [PITH_FULL_IMAGE… view at source ↗
Figure 2
Figure 2. The dotted vertical lines in the background mark transitions between passes. As expected, [PITH_FULL_IMAGE:figures/full_fig_p045_2.png] view at source ↗
Figure 3
Figure 3. All curves are plotted for 20 passes. The linear extrapolation of each curve onto the [PITH_FULL_IMAGE:figures/full_fig_p046_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: The correlation with the previous spin value at zero temperature. [PITH_FULL_IMAGE:figures/full_fig_p047_4.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Markov Chains Approximate Message Passing

    cs.DS 2025-12 conditional novelty 6.0 of 10

    For spiked Wigner inference, Glauber dynamics and AMP reach the same correlation fixed point, with a phase transition at βλ=1 conditional on SK mixing.

Reference graph

Works this paper leans on

53 extracted references · 41 canonical work pages · cited by 1 Pith paper

  1. [1]

    Aging of spherical spin glasses

    G Ben Arous, Amir Dembo, and Alice Guionnet. Aging of spherical spin glasses. Probability theory and related fields , 120:1--67, 2001

    work page 2001

  2. [2]

    Large deviations for langevin spin glass dynamics

    G Ben Arous and Alice Guionnet. Large deviations for langevin spin glass dynamics. Probability Theory and Related Fields , 102:455--509, 1995

    work page 1995

  3. [3]

    Symmetric langevin spin glass dynamics

    G Ben Arous and Alice Guionnet. Symmetric langevin spin glass dynamics. The Annals of Probability , 25(3):1367--1422, 1997

    work page 1997

  4. [4]

    Near-optimal shattering in the Ising pure p-spin and rarity of solutions returned by stable algorithms

    Ahmed El Alaoui. Near-optimal shattering in the ising pure p-spin and rarity of solutions returned by stable algorithms. arXiv preprint arXiv:2412.03511 , 2024

    work page Pith review arXiv 2024

  5. [5]

    Sampling from mean-field gibbs measures via diffusion processes

    Ahmed El Alaoui, Andrea Montanari, and Mark Sellke. Sampling from mean-field gibbs measures via diffusion processes. arXiv preprint arXiv:2310.08912 , 2023

    arXiv 2023

  6. [6]

    Aging and spin-glass dynamics

    G \'e rard Ben-Arous. Aging and spin-glass dynamics. arXiv preprint math/0304364 , 2003

    work page Pith review arXiv 2003

  7. [7]

    Cugliandolo-Kurchan equations for dynamics of spin-glasses

    G \'e rard Ben Arous, Amir Dembo, and Alice Guionnet. Cugliandolo-Kurchan equations for dynamics of spin-glasses. Probability theory and related fields , 136(4):619--660, 2006

    work page 2006

  8. [8]

    Energy-decreasing dynamics in mean-field spin models

    L Bussolari, P Contucci, M Degli Esposti, and Cristian Giardina. Energy-decreasing dynamics in mean-field spin models. Journal of Physics A: Mathematical and General , 36(10):2413, 2003

    work page 2003

Show all 53 references
  1. [9]

    The dynamics of message passing on dense graphs, with applications to compressed sensing

    Mohsen Bayati and Andrea Montanari. The dynamics of message passing on dense graphs, with applications to compressed sensing. IEEE Transactions on Information Theory , 57(2):764--785, 2011

    2011

  2. [10]

    An iterative construction of solutions of the tap equations for the Sherrington--Kirkpatrick model

    Erwin Bolthausen. An iterative construction of solutions of the tap equations for the Sherrington--Kirkpatrick model. Communications in Mathematical Physics , 325(1):333--366, 2014

    2014

  3. [11]

    The high-dimensional asymptotics of first order methods with random data

    Michael Celentano, Chen Cheng, and Andrea Montanari. The high-dimensional asymptotics of first order methods with random data. arXiv preprint arXiv:2112.07572 , 2021

    2021 arXiv

  4. [12]

    Interpolating greedy and reluctant algorithms

    Pierluigi Contucci, Cristian Giardin \`a , Claudio Giberti, Francesco Unguendoli, and Cecilia Vernia. Interpolating greedy and reluctant algorithms. Optimization Methods and Software , 20(4-5):509--514, 2005

    2005

  5. [13]

    Finding minima in complex landscapes: annealed, greedy and reluctant algorithms

    Pierluigi Contucci, Cristian Giardina, Claudio Giberti, and Cecilia Vernia. Finding minima in complex landscapes: annealed, greedy and reluctant algorithms. Mathematical Models and Methods in Applied Sciences , 15(09):1349--1369, 2005

    2005

  6. [14]

    Analytical solution of the off-equilibrium dynamics of a long-range spin-glass model

    Leticia F Cugliandolo and Jorge Kurchan. Analytical solution of the off-equilibrium dynamics of a long-range spin-glass model. Physical Review Letters , 71(1):173, 1993

    1993

  7. [15]

    On the out-of-equilibrium relaxation of the Sherrington-Kirkpatrick model

    Leticia F Cugliandolo and Jorge Kurchan. On the out-of-equilibrium relaxation of the Sherrington-Kirkpatrick model. Journal of Physics A: Mathematical and General , 27(17):5749, 1994

    1994

  8. [16]

    Almost-linear planted cliques elude the metropolis process

    Zongchen Chen, Elchanan Mossel, and Ilias Zadik. Almost-linear planted cliques elude the metropolis process. Random Structures & Algorithms , 66(2):e21274, 2025

    2025

  9. [17]

    Analysis of the infinite-replica symmetry breaking solution of the Sherrington-Kirkpatrick model

    Andrea Crisanti and Tommaso Rizzo. Analysis of the infinite-replica symmetry breaking solution of the Sherrington-Kirkpatrick model. Physical Review E , 65(4):046137, 2002

    2002

  10. [18]

    Systematic scan for sampling colorings

    Martin Dyer, Leslie Ann Goldberg, and Mark Jerrum. Systematic scan for sampling colorings. The Annals of Applied Probability , 16(1):185--230, 2006

    2006

  11. [19]

    Dobrushin conditions and systematic scan

    Martin Dyer, Leslie Ann Goldberg, and Mark Jerrum. Dobrushin conditions and systematic scan. Combinatorics, Probability and Computing , 17(6):761--779, 2008

    2008

  12. [20]

    Theory of spin glasses

    Samuel Frederick Edwards and Phil W Anderson. Theory of spin glasses. Journal of Physics F: Metal Physics , 5(5):965, 1975

    1975

  13. [21]

    Quenches in the sherrington--kirkpatrick model

    Vittorio Erba, Freya Behrens, Florent Krzakala, and Lenka Zdeborov \'a . Quenches in the sherrington--kirkpatrick model. Journal of Statistical Mechanics: Theory and Experiment , 2024(8):083302, 2024

    2024

  14. [22]

    A spectral condition for spectral gap: fast mixing in high-temperature ising models

    Ronen Eldan, Frederic Koehler, and Ofer Zeitouni. A spectral condition for spectral gap: fast mixing in high-temperature ising models. Probability theory and related fields , 182(3):1035--1051, 2022

    2022

  15. [23]

    New method for studying the dynamics of disordered spin systems without finite-size effects

    H Eissfeller and M Opper. New method for studying the dynamics of disordered spin systems without finite-size effects. Physical review letters , 68(13):2094, 1992

    1992

  16. [24]

    Mean-field monte carlo approach to the Sherrington-Kirkpatrick model with asymmetric couplings

    H Eissfeller and M Opper. Mean-field monte carlo approach to the Sherrington-Kirkpatrick model with asymmetric couplings. Physical Review E , 50(2):709, 1994

    1994

  17. [25]

    Multi-species asymmetric exclusion process in ordered sequential update

    ME Fouladvand and F Jafarpour. Multi-species asymmetric exclusion process in ordered sequential update. Journal of Physics A: Mathematical and General , 32(32):5845, 1999

    1999

  18. [26]

    Real analysis: modern techniques and their applications

    Gerald B Folland. Real analysis: modern techniques and their applications . John Wiley & Sons, 1999

    1999

  19. [27]

    A unifying tutorial on approximate message passing

    Oliver Y Feng, Ramji Venkataramanan, Cynthia Rush, Richard J Samworth, et al. A unifying tutorial on approximate message passing. Foundations and Trends in Machine Learning , 15(4):335--536, 2022

    2022

  20. [28]

    Shattering in the ising pure p -spin model

    David Gamarnik, Aukosh Jagannath, and Eren C K z lda g . Shattering in the ising pure p -spin model. arXiv preprint arXiv:2307.07461 , 2023

    2023 arXiv

  21. [29]

    Hardness of random optimization problems for boolean circuits, low-degree polynomials, and langevin dynamics

    David Gamarnik, Aukosh Jagannath, and Alexander S Wein. Hardness of random optimization problems for boolean circuits, low-degree polynomials, and langevin dynamics. SIAM Journal on Computing , 53(1):1--46, 2024

    2024

  22. [30]

    Sanov results for glauber spin-glass dynamics

    Malte Grunwald. Sanov results for glauber spin-glass dynamics. Probability theory and related fields , 106(2):187--232, 1996

    1996

  23. [31]

    Rigorous dynamical mean-field theory for stochastic gradient descent methods

    Cedric Gerbelot, Emanuele Troiani, Francesca Mignacco, Florent Krzakala, and Lenka Zdeborova. Rigorous dynamical mean-field theory for stochastic gradient descent methods. SIAM Journal on Mathematics of Data Science , 6(2):400--427, 2024

    2024

  24. [32]

    Broken replica symmetry bounds in the mean field spin glass model

    Francesco Guerra. Broken replica symmetry bounds in the mean field spin glass model. Communications in mathematical physics , 233:1--12, 2003

    2003

  25. [33]

    Averaged and quenched propagation of chaos for spin glass dynamics

    Alice Guionnet. Averaged and quenched propagation of chaos for spin glass dynamics. Probability Theory and Related Fields , 109:183--215, 1997

    1997

  26. [34]

    Sampling from spherical spin glasses in total variation via algorithmic stochastic localization

    Brice Huang, Andrea Montanari, and Huy Tuan Pham. Sampling from spherical spin glasses in total variation via algorithmic stochastic localization. arXiv preprint arXiv:2404.15651 , 2024

    2024 arXiv

  27. [35]

    Weak poincaré inequalities, simulated annealing, and sampling from spherical spin glasses

    Brice Huang, Sidhanth Mohanty, Amit Rajaraman, and David X Wu. Weak poincaré inequalities, simulated annealing, and sampling from spherical spin glasses. arXiv preprint arXiv:2411.09075 , 2024

    2024 arXiv

  28. [36]

    Locally stationary distributions: A framework for analyzing slow-mixing markov chains

    Kuikui Liu, Sidhanth Mohanty, Prasad Raghavendra, Amit Rajaraman, and David X Wu. Locally stationary distributions: A framework for analyzing slow-mixing markov chains. In 2024 IEEE 65th Annual Symposium on Foundations of Computer Science (FOCS) , pages 203--215. IEEE, 2024

    2024

  29. [37]

    Private communication

    James MacLaurin. Private communication

  30. [38]

    An emergent autonomous flow for mean-field spin glasses

    James MacLaurin. An emergent autonomous flow for mean-field spin glasses. Probability Theory and Related Fields , 180(1):365--438, 2021

    2021

  31. [39]

    Optimization of random cost functions and statistical physics

    Andrea Montanari. Optimization of random cost functions and statistical physics. arXiv preprint arXiv:2401.11348 , 2024

    2024 arXiv

  32. [40]

    The Parisi ultrametricity conjecture

    Dmitry Panchenko. The Parisi ultrametricity conjecture. Annals of Mathematics , pages 383--393, 2013

    2013

  33. [41]

    The Sherrington-Kirkpatrick model

    Dmitry Panchenko. The Sherrington-Kirkpatrick model . Springer Science & Business Media, 2013

    2013

  34. [42]

    Infinite number of order parameters for spin-glasses

    Giorgio Parisi. Infinite number of order parameters for spin-glasses. Physical Review Letters , 43(23):1754, 1979

    1979

  35. [43]

    The order parameter for spin glasses: a function on the interval 0-1

    Giorgio Parisi. The order parameter for spin glasses: a function on the interval 0-1. Journal of Physics A: Mathematical and General , 13(3):1101, 1980

    1980

  36. [44]

    A sequence of approximated solutions to the sk model for spin glasses

    Giorgio Parisi. A sequence of approximated solutions to the sk model for spin glasses. Journal of Physics A: Mathematical and General , 13(4):L115, 1980

    1980

  37. [45]

    On the statistical properties of the large time zero temperature dynamics of the sk model

    Giorgio Parisi. On the statistical properties of the large time zero temperature dynamics of the sk model. Fractals , 11(supp01):161--171, 2003

    2003

  38. [46]

    Can extra updates delay mixing? Communications in Mathematical Physics , 323:1007--1016, 2013

    Yuval Peres and Peter Winkler. Can extra updates delay mixing? Communications in Mathematical Physics , 323:1007--1016, 2013

    2013

  39. [47]

    Sompolinsky, A

    H. Sompolinsky, A. Crisanti, and H. J. Sommers. Chaos in random neural networks. Phys. Rev. Lett. , 61:259--262, Jul 1988

    1988

  40. [48]

    The threshold energy of low temperature langevin dynamics for pure spherical spin glasses

    Mark Sellke. The threshold energy of low temperature langevin dynamics for pure spherical spin glasses. Communications on Pure and Applied Mathematics , 77(11):4065--4099, 2024

    2024

  41. [49]

    Approximate counting, uniform generation and rapidly mixing markov chains

    Alistair Sinclair and Mark Jerrum. Approximate counting, uniform generation and rapidly mixing markov chains. Information and Computation , 82(1):93--133, 1989

    1989

  42. [50]

    Solvable model of a spin-glass

    David Sherrington and Scott Kirkpatrick. Solvable model of a spin-glass. Physical review letters , 35(26):1792, 1975

    1975

  43. [51]

    Relaxational dynamics of the Edwards-Anderson model and the mean-field theory of spin-glasses

    Haim Sompolinsky and Annette Zippelius. Relaxational dynamics of the Edwards-Anderson model and the mean-field theory of spin-glasses. Physical Review B , 25(11):6860, 1982

    1982

  44. [52]

    The Parisi formula

    Michel Talagrand. The Parisi formula. Annals of mathematics , pages 221--263, 2006

    2006

  45. [53]

    Solution of'solvable model of a spin glass'

    David J Thouless, Philip W Anderson, and Robert G Palmer. Solution of'solvable model of a spin glass'. Philosophical Magazine , 35(3):593--601, 1977

    1977

Pith tools

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