pith. sign in

arxiv: 2606.18752 · v2 · pith:RDJSPQNAnew · submitted 2026-06-17 · 🧮 math-ph · cond-mat.dis-nn· math.MP

Self-averaging of replica overlaps in the random field Edwards-Anderson model

C. Itoi , Y. Sakamoto This is my paper

Pith reviewed 2026-06-26 19:19 UTC · model grok-4.3

classification 🧮 math-ph cond-mat.dis-nnmath.MP
keywords Edwards-Anderson modelrandom fieldself-averagingreplica overlapspin glassTasaki correlation inequalityinfinite volume limitfree energy derivative
0
0 comments X

The pith

The replica overlap self-averages in the random-field Edwards-Anderson model almost everywhere in coupling space.

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

The paper proves that the replica overlap self-averages in the Edwards-Anderson model with random fields. This means the variance of the overlap goes to zero in the infinite volume limit almost everywhere in the space of coupling constants. The result holds in any dimension. It is shown using a representation of the order parameter as the derivative of the free energy with respect to the random field strength and Tasaki's correlation inequality that bounds the overlap fluctuations. A parallel proof applies to the bond overlap in the Gaussian interaction case without random field.

Core claim

The self-averaging of the replica overlap is proven in the Edwards-Anderson (EA) model under random field almost everywhere in the coupling constant space in any dimension. The variance of the replica overlap vanishes in the infinite-volume limit. The EA order parameter is represented in terms of the derivative of the free energy density with respect to the random field strength, regardless of boundary conditions. Tasaki's correlation inequality for finite-dimensional spin glass models shows that the expectation of the squared replica overlap is bounded by the squared EA order parameter. These simple evaluations enable us to prove that the variance of the replica overlap vanishes in the infi

What carries the argument

Representation of the EA order parameter as the derivative of the free energy density with respect to random field strength, used with Tasaki's correlation inequality to bound the replica overlap.

Load-bearing premise

Tasaki's correlation inequality applies to the finite-dimensional spin glass models considered, and the representation of the EA order parameter as the derivative of the free energy density holds regardless of boundary conditions.

What would settle it

Finding a specific value of the coupling constants where the variance of the replica overlap does not approach zero as the system size goes to infinity while the order parameter remains positive.

Figures

Figures reproduced from arXiv: 2606.18752 by C. Itoi, Y. Sakamoto.

Figure 1
Figure 1. Figure 1: Black dots and white dots represent the sites in Λ [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Four translated copies of Λ3 are embedded into Λ9. This corresponds to the optimal choice (4.35). Fix any κ 6= κ ′ and abbreviate Λκ ℓ and Λκ ′ ℓ as Λa and Λb, respectively. Decompose any spin configuration σ = (σx)x∈ΛL ∈ CL into σ = (σa, σb, τ), where σa = (σx)x∈Λa , σb = (σx)x∈Λb , and τ = (τx)x∈ΛL\(Λa∪Λb) . Here, we denote τx = σx at x ∈ ΛL\(Λa ∪ Λb) for later convenience. Let us tentatively fix a bound… view at source ↗
read the original abstract

The self-averaging of the replica overlap is proven in the Edwards-Anderson (EA) model under random field almost everywhere in the coupling constant space in any dimension. The EA order parameter is represented in terms of the derivative of the free energy density with respect to the random field strength, regardless of boundary conditions. Tasaki's correlation inequality for finite-dimensional spin glass models shows that the expectation of the squared replica overlap is bounded by the squared EA order parameter. These simple evaluations enable us to prove that the variance of the replica overlap vanishes in the infinite-volume limit. The self-averaging of the replica bond overlap is proven also in the EA model with Gaussian exchange interaction without random field. Short-range spin glass models have been shown to behave differently from mean-field spin glass models with RSB phase.

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 / 1 minor

Summary. The paper proves self-averaging of the replica overlap in the random-field Edwards-Anderson model almost everywhere in the coupling-constant space, in any dimension. The EA order parameter is represented as the derivative of the quenched free-energy density with respect to random-field strength (independent of boundary conditions). Tasaki's correlation inequality is invoked to bound the expectation of the squared replica overlap by the square of this order parameter. Together these yield that the variance of the overlap vanishes in the infinite-volume limit. Self-averaging of the replica bond overlap is also shown for the Gaussian-exchange EA model without random field.

Significance. If the arguments hold, the result is significant: it supplies a rigorous, inequality-based proof that short-range spin glasses self-average their overlaps, in contrast to mean-field models with RSB. The derivation is parameter-free once the free-energy representation and Tasaki bound are granted, and it applies in arbitrary dimension, which strengthens the distinction between finite-dimensional and mean-field behavior.

major comments (2)
  1. [main proof (following the free-energy representation of q_EA)] The central step invokes Tasaki's correlation inequality to obtain E[(replica overlap)²] ≤ q_EA² for the random-field Hamiltonian. Tasaki's original statement applies to finite-dimensional spin-glass models without an explicit random-field term; the manuscript must supply an explicit verification (or a cited extension) that the inequality remains valid once the random-field term is added and for the boundary conditions used in the infinite-volume limit. This bound is load-bearing for the variance-vanishing claim.
  2. [section introducing the free-energy representation] The representation of q_EA as the derivative of the free-energy density with respect to random-field strength is asserted to hold regardless of boundary conditions. A self-contained derivation showing that the infinite-volume limit of this derivative coincides with the EA order parameter for the chosen boundary conditions (periodic, free, etc.) is needed, because any dependence on boundary conditions would affect the subsequent application of Tasaki's inequality.
minor comments (1)
  1. [abstract and main theorem statement] The phrase 'almost everywhere in the coupling constant space' should be accompanied by a brief statement of the measure with respect to which the exceptional set has zero measure (Lebesgue, etc.).

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments. Below we respond point by point to the major comments. We will revise the manuscript to supply the requested verifications and derivations.

read point-by-point responses

  1. Referee: [main proof (following the free-energy representation of q_EA)] The central step invokes Tasaki's correlation inequality to obtain E[(replica overlap)²] ≤ q_EA² for the random-field Hamiltonian. Tasaki's original statement applies to finite-dimensional spin-glass models without an explicit random-field term; the manuscript must supply an explicit verification (or a cited extension) that the inequality remains valid once the random-field term is added and for the boundary conditions used in the infinite-volume limit. This bound is load-bearing for the variance-vanishing claim.

    Authors: We agree that explicit verification is required. Tasaki's inequality is formulated in terms of the spin-spin interaction terms on the lattice graph; the random-field term is an additive external field linear in the individual spins and does not modify the interaction structure or the sign conditions used in the proof of the inequality. Consequently the same bound applies verbatim to the random-field Hamiltonian. In the revised manuscript we will insert a short appendix paragraph that recalls the precise hypotheses of Tasaki's theorem and verifies that they continue to hold when the random-field term is present, for both periodic and free boundary conditions in the infinite-volume limit. revision: yes

  2. Referee: [section introducing the free-energy representation] The representation of q_EA as the derivative of the free-energy density with respect to random-field strength is asserted to hold regardless of boundary conditions. A self-contained derivation showing that the infinite-volume limit of this derivative coincides with the EA order parameter for the chosen boundary conditions (periodic, free, etc.) is needed, because any dependence on boundary conditions would affect the subsequent application of Tasaki's inequality.

    Authors: The finite-volume overlap is exactly the derivative of the finite-volume free energy with respect to the random-field strength. The quenched free-energy density converges to a limit that is known to be independent of boundary conditions for the EA model (by standard subadditive arguments). Differentiation and the thermodynamic limit may be interchanged because the derivatives are uniformly bounded. We will add a concise, self-contained derivation of this interchange in the revised text, making explicit that the resulting infinite-volume order parameter is the same for the boundary conditions under consideration and therefore compatible with the subsequent application of Tasaki's inequality. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation relies on external inequality and standard representation

full rationale

The paper's central argument represents the EA order parameter as the derivative of the quenched free-energy density with respect to random-field strength (a standard thermodynamic identity independent of the target variance result) and invokes Tasaki's correlation inequality (an external result from prior literature by a different author) to bound E[(replica overlap)^2] by the squared order parameter. These steps do not reduce the variance vanishing claim to a self-definition, a fitted input renamed as prediction, or a load-bearing self-citation chain; the proof is self-contained against external mathematical benchmarks and contains no ansatz smuggling or renaming of known empirical patterns.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The proof rests on standard mathematical assumptions in statistical mechanics and a specific correlation inequality.

axioms (2)
  • domain assumption Tasaki's correlation inequality holds for the models considered
    Used to bound the expectation of squared replica overlap.
  • domain assumption The EA order parameter can be represented as derivative of free energy density w.r.t. random field strength
    Regardless of boundary conditions, as stated.

pith-pipeline@v0.9.1-grok · 5670 in / 1104 out tokens · 28343 ms · 2026-06-26T19:19:09.623432+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

32 extracted references · 9 linked inside Pith

  1. [1]

    On properties of Parisi measures

    Auffinger, A., Chen, W-K. On properties of Parisi measures. Probab. Theory. Relat. Fields, Vol. 161 pp. 817-850, 2015. \ ://arxiv.org/abs/1303.3573

    Pith/arXiv arXiv 2015

  2. [2]

    On the stability of quenched state in mean-field spin glass models

    Aizenman, M., Contucci, P. On the stability of quenched state in mean-field spin glass models. J. Stat. Phys. Vol. 92, pp. 765-783, 1997. \ ://arxiv.org/abs/cond-mat/9712129

    Pith/arXiv arXiv 1997

  3. [3]

    Almeida, de J. R. L. and D. J. Thouless. Stability of the Sherrington-Kirkpatrick solution of a spin glass model, J. Phys. A Math. Gen. , Vol. 11, No. 5, pp. 983--990, 1978

    1978

  4. [4]

    Statistical Mechanics of Disordered Systems: A Mathematical Perspective , Cambridge Series in Statistical and Probabilistic Mathematics, Cambridge University Press, 2006

    Bovier, A. Statistical Mechanics of Disordered Systems: A Mathematical Perspective , Cambridge Series in Statistical and Probabilistic Mathematics, Cambridge University Press, 2006

    2006

  5. [5]

    Br\'ezin, E., De Dominicis, C., New phenomena in the random field Ising model, Euro. Phys. Lett. Vol. 44, No. 1, pp.13--19, 1998

    1998

  6. [6]

    Absence of Replica Symmetry Breaking in the Random Field Ising Model, Commun

    Chatterjee, S. Absence of Replica Symmetry Breaking in the Random Field Ising Model, Commun. Math. Phys. , Vol. 337, pp. 93--102, 2015. \ ://arxiv.org/abs/1404.7178

    Pith/arXiv arXiv 2015

  7. [7]

    Chen, W-K., On the Almeida-Thouless transition line in the Sherrington-Kirkpatrick model with centered Gaussian external field, Electron Commun. Probab. Vol. 26, No. 65, pp. 1-9. \ ://arxiv.org/abs/2103.04802

    arXiv

  8. [8]

    Edwards, S. F. and P. W. Anderson. Theory of spin glasses, J. Phys. F. , Vol. 5, No. 5, p. 965, 1975

    1975

  9. [9]

    Enter, van A. C. D. and R. B. Griffiths. The order parameter in a spin glass, Commun. Math. Phys. , Vol. 90, No. 3, pp. 319--327, 1983. \ ://projecteuclid.org/journals/communications-in-mathematical-physics/volume-90/issue-3/The-order-parameter-in-a-spin-glass/cmp/1103940344.full

    arXiv 1983

  10. [10]

    Fisher, D. S. and D. A. Huse. Ordered Phase of Short-Range Ising Spin-Glasses, Phys. Rev. Lett. , Vol. 56, pp. 1601--1604, 1986

    1986

  11. [11]

    Fisher, D. S. and D. A. Huse. Equilibrium behavior of the spin-glass ordered phase, Phys. Rev. B , Vol. 38, pp. 386--411, 1988

    1988

  12. [12]

    Friedli, S. and Y. Velenik. Statistical Mechanics of Lattice Systems: A Concrete Mathematical Introduction , Cambridge University Press, 2017

    2017

  13. [13]

    M., Kasteleyn P

    Fortuin,C. M., Kasteleyn P. W., Ginibre, J. Correlation inequalities on some partially ordered sets.Commun. Math. Phys. Vol. 22, pp. 89-103, 1971

    1971

  14. [14]

    Griffiths, R. B. Spontaneous Magnetization in Idealized Ferromagnets, Phys. Rev. , Vol. 152, pp. 240--246, 1966

    1966

  15. [15]

    : General properties of overlap probability distributions in disordered spin systems

    Ghirlanda, S., Guerra, F. : General properties of overlap probability distributions in disordered spin systems. Towards Parisi ultrametricity . J. Phys. A, Vol. 31, pp. 9149--9155, 1998

    1998

  16. [16]

    Mukaida, H

    Itoi, C. Mukaida, H. and Tasaki, H. Griffiths-Type Theorems for Short-Range Spin Glass Models, Vol. 191, No. 28, pp.1-30, 2024. \ ://arxiv.org/abs/2310.04775

    arXiv 2024

  17. [17]

    Koma, T. and H. Tasaki. Symmetry breaking in Heisenberg antiferromagnets, Commun. Math. Phys , Vol. 158, No. 1, pp. 191--214, 1993. \ ://projecteuclid.org/journals/communications-in-mathematical-physics/volume-158/issue-1/Symmetry-breaking-in-Heisenberg-antiferromagnets/cmp/1104254136.full

    arXiv 1993

  18. [18]

    and Martin, C., Spin and link overlaps in three-dimensional spin glasses, Phys

    Krzakala, F. and Martin, C., Spin and link overlaps in three-dimensional spin glasses, Phys. Rev. Lett. Vol. 85, No. 14, pp. 3013--3016, 2000. \ ://arxiv.org/abs/cond-mat/0002055

    Pith/arXiv arXiv 2000

  19. [19]

    and Young, A.P., Replica symmetry breaking in the random field Ising model, Euro

    M\'ezard, M. and Young, A.P., Replica symmetry breaking in the random field Ising model, Euro. Phys. Lett. Vol. 18, No. 7, pp 653--659 . 1992. ://arxiv.org/abs/cond-mat/9804266

    Pith/arXiv arXiv 1992

  20. [20]

    Newman, C. M. and D. L. Stein. Non-Mean-Field Behavior of Realistic Spin Glasses, Phys. Rev. Lett. , Vol. 76, pp. 515--518, 1996. \ ://arxiv.org/abs/adap-org/9508006

    Pith/arXiv arXiv 1996

  21. [21]

    Newman, C. M. and D. L. Stein. Metastate approach to thermodynamic chaos, Phys. Rev. E , Vol. 55, pp. 5194--5211, 1997. \ ://arxiv.org/abs/cond-mat/9612097

    Pith/arXiv arXiv 1997

  22. [22]

    Newman, C. M. and D. L. Stein. The state(s) of replica symmetry breaking: Mean field theories vs. short-ranged spin glasses, J. Stat. Phys , Vol. 106, pp. 213--244, 2002. \ ://arxiv.org/abs/cond-mat/0105282

    Pith/arXiv arXiv 2002

  23. [23]

    Statistical Physics of Spin Glasses and Information

    Nishimori, H. Statistical Physics of Spin Glasses and Information

  24. [24]

    Nishimori,H. and D. Sherrington, D. Absence of replica symmetry breaking in a region

  25. [25]

    Infinite Number of Order Parameters for Spin-Glasses, Phys

    Parisi, G. Infinite Number of Order Parameters for Spin-Glasses, Phys. Rev. Lett. , Vol. 43, pp. 1754--1756, 1979

    1979

  26. [26]

    The Parisi formula for mixed p -spin models , Ann

    Panchenko, D. The Parisi formula for mixed p -spin models , Ann. Probab. , Vol. 42, No. 3, pp. 946 -- 958, 2014. \ ://arxiv.org/abs/1112.4409

    Pith/arXiv arXiv 2014

  27. [27]

    Sherrington, D. and S. Kirkpatrick. Solvable Model of a Spin-Glass, Phys. Rev. Lett. , Vol. 35, pp. 1792--1796, 1975

    1975

  28. [28]

    The statistical mechanics of lattice gases, Vol

    Simon, B. The statistical mechanics of lattice gases, Vol. 1 , Princeton Series in Physics, Princeton University Press, 1993

    1993

  29. [29]

    Spin Glasses

    Talagrand, M. Spin Glasses. A Challenge for Mathematicians: Cavity and Mean Field Models , Springer, 2003

    2003

  30. [30]

    The Parisi formula, Ann

    Talagrand, M. The Parisi formula, Ann. Math. , Vol. 163, No. 1, pp. 221--263, 2006

    2006

  31. [31]

    Mean Field Models for Spin Glasses , Vol

    Talagrand, M. Mean Field Models for Spin Glasses , Vol. I and II Springer, 2011

    2011

  32. [32]

    On the upper critical dimensions of random spin systems, J

    Tasaki, H. On the upper critical dimensions of random spin systems, J. Stat. Phys. , Vol. 54, No. 1, pp. 163--170, 1989

    1989