Pith. sign in

REVIEW 4 major objections 4 minor 60 references

Guerra interpolation for inverse freezing

T0 review · 4 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A unified Hamiltonian plus Guerra interpolation gives closed-form free energies for the main inverse-freezing spin-glass models, recovering replica-trick results and adding a new instability line.

desk verdict A useful interpolation-method paper with a sign error in the K-coupling that invalidates the unified and BEGC/AT-line formulas; the K=0 re-derivations are fine. read the letter →

arxiv 2505.06202 v1 pith:5XWWVS2Q submitted 2025-05-09 cond-mat.dis-nn cond-mat.stat-mech

Linda Albanese , Adriano Barra , Emilio N. M. Cirillo This is my paper · ORCID
classification cond-mat.dis-nncond-mat.stat-mech
keywords GuerrainterpolationinversefreezingspinglassGhatak-SherringtonmodelreplicasymmetrybreakingdeAlmeida-ThoulesslinequenchedfreeenergyBlume-Emery-Griffiths-Capel
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 tries to give one derivation, not a pile of separate calculations, for the equilibrium free energies of the main mean-field spin-glass models that show inverse freezing. It writes a single Hamiltonian with multi-valued spins, a crystal-field term $D\sum_i s_i^2$, and a biquadratic coupling $K\sum_{i

What carries the argument

The device that carries the argument is an interpolating statistical pressure $A_{N,D,K,J,J_0}(t|\beta)$ built from partition function (4.2): at $t=1$ it is the original model, at $t=0$ it is a one-body model whose pressure is computable, and the Fundamental Theorem of Calculus connects the two through the $t$-derivative. Under Assumption 1 the derivative simplifies in the thermodynamic limit to the $t$-independent constant (4.33), with the interpolation parameters fixed by (4.39) as $\psi=\beta J_0\bar m$, $A^2=\beta^2J^2\bar q$, and $B=\beta^2J^2(Q-\bar q)-2\beta D-2\beta KQ$, so that the order-parameter variances vanish. For 1-RSB, the same scheme is run with two Gaussian fields and the hierarchical averaging (4.44)-(4.46), and this yields Theorem 2.

What would settle it

Run a Monte Carlo simulation of the Ghatak-Sherrington model at parameters inside the inverse-freezing region, and compare the simulated quenched pressure with the RS formula (4.16) at increasing $N$; if the difference does not go to zero, Theorem 1 fails there. Equally direct: measure the finite-size variance of $q_{ab}$, $q_{aa}$, and $m$; if the variance does not vanish as $N\to\infty$, Assumption 1, and with it the derivation, is violated.

Watch

Extended reading notes

Core claim

The central claim is that the replica-symmetric quenched statistical pressure of Hamiltonian (2.1) is, in the thermodynamic limit, the explicit formula $$\mathbb{E}\log\sum_{s\in\$\Omega$}\exp[\$\beta$(h+J_0\bar m+J\sqrt{\bar q}z)s + \$\beta$(\frac{\$\beta$ $J^{2}$}{2}(Q-\bar q)-D+KQ)$s^{2}$] - \frac{\$\beta$ J_0}{2}\bar $m^{2}$ - \frac{\$beta^{2}$$J^{2}$}{4}($Q^{2}$-\bar $q^{2}$) - \frac{\$\beta$ K}{2}$Q^{2}$,$$ where $\mathbb{E}$ averages over the external field $h$ and the standard Gaussian $z$, and the order parameters $\bar m, Q, \bar q$ solve the self-consistency equations (4.8)-(4.10). Theorem 2 extends this to one-step replica symmetry breaking with two overlap values, and Theorem 3 gives the instability boundary below which the RS solution is not the stable description. The proof route is Guerra interpolation rather than the replica trick, and the corollaries are claimed to agree with earlier heuristic calculations.

Load-bearing premise

The load-bearing premise is that, in the thermodynamic limit, the order parameters concentrate: the magnetization $m$, the overlap $q_{ab}$, and the self-overlap $q_{aa}$ all become deterministic functions of their mean values, as stated in Assumption 1. If that concentration fails, the interpolation derivative does not simplify to (4.33), and the closed-form free energies are not established; the paper also assumes the thermodynamic limit of the quenched free energy exists (Remark 3).

Editorial extensions

If this is right

  • Theorem 1's RS formula gives, in one expression, the quenched free energy of the Ghatak-Sherrington, Katayama-Horiguchi, and disordered Blume-Emery-Griffiths-Capel mean-field models in their replica-symmetric regimes.
  • The 1-RSB formula (Theorem 2) covers the symmetry-broken phase and reduces exactly to the RS result when the parameter $\theta$ goes to 0 or 1.
  • The general AT line (4.92) recovers the known Sherrington-Kirkpatrick and Ghatak-Sherrington stability thresholds and yields, for the disordered BEGC mean-field model, a threshold not previously recorded.
  • The annealed pressures, some of which are new, provide upper bounds for the quenched free energies by Jensen's inequality.
  • Because the unified Hamiltonian contains all submodels as limits, the same derivation chain covers all of them at once.

Reading between the lines

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

  • Going beyond the paper, the same interpolation with the generalized Hamiltonian could be applied to multi-state neural-network variants, such as Blume-Capel Hopfield-like architectures, to test whether inverse-freezing-like transitions appear there.
  • Going beyond the paper, the general AT line (4.92) can be made explicit for each submodel and checked numerically: where it crosses the inverse-freezing transition boundary, the RS description should be abandoned in favor of the 1-RSB free energy.
  • Going beyond the paper, a Monte Carlo measurement of the finite-size variance of $q_{ab}$, $q_{aa}$, and $m$ would test whether the self-averaging assumption holds exactly in the region where inverse freezing occurs; if it fails, the RS formula should be read as an upper bound rather than the exact pressure.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The manuscript introduces a unified mean-field Hamiltonian (2.1) with spins s ∈ {−1+γ/S}, Gaussian couplings, crystal field D, biquadratic coupling K, and random external field h. It derives the annealed pressure and, via Guerra interpolation, the replica-symmetric and first-step replica-symmetry-broken quenched pressures (Theorems 1 and 2), together with self-consistency equations and an Almeida-Thouless instability line (Theorem 3). It then checks the Sherrington-Kirkpatrick, Ghatak-Sherrington, Katayama-Horiguchi, and disordered Blume-Emery-Griffiths-Capel limits, claiming agreement with known replica results and presenting some expressions it identifies as new.

Significance. If correct, the paper would provide a unified interpolation-based derivation of several known inverse-freezing spin-glass free energies and a new AT-line formula. Strengths include the self-contained character of the derivations, the recovery of known K=0 limits (SK, GS, KH) without importing replica results, and the AT-line derivation from the paper's own 1-RSB expression rather than from a fitted ansatz. However, the K-dependent statements are not internally consistent: the sign mismatch between the stated pressures and their stationary equations means the unified-model and BEGC results cannot be used as printed. Because the inconsistencies are finite-N algebra issues rather than conceptual dead ends, they are fixable within the manuscript's scope, but they must be corrected before the central claims can be accepted.

major comments (4)
  1. [§4.1, Theorem 1 (Eq. (4.7)) and Eqs. (4.8)–(4.10); Lemma 2, Eq. (4.33) vs (4.38)] The RS pressure in (4.7) contains +βKQ in the single-site exponent and −βKQ²/2 outside, but the self-consistency equations (4.8)–(4.10) contain −βKQ in the same exponent. Differentiating (4.7) with respect to Q yields a stationary condition with +βKQ, so (4.8)–(4.10) do not extremize (4.7) when K ≠ 0. The same +/− flip appears in Remark 7, in Corollary 12, and in the 1-RSB formulas in §4.2. Moreover, Lemma 2 as stated in (4.33) has −βKQ²/2, while the proof in (4.38) produces +βKQ²/2. These inconsistencies make every K-dependent formula in the paper unusable until the sign convention is fixed.
  2. [§4.2, Theorem 2 (Eq. (4.54)) and Corollary 18 (Eqs. (4.79)–(4.82))] The same mismatch appears at the 1-RSB level: the pressure (4.54) has −βKQ in the single-site exponent, while the weight W in the self-consistency equations printed after (4.58) has +βKQ; similarly, the pressure in (4.79) has −βKQ while the ξ used in (4.80)–(4.82) is +βKQ. Remark 13's reduction of (4.54) to the RS expression ends with −βKQ in the exponent, in conflict with (4.7). The likely origin is the treatment of the biquadratic term: (2.1) uses −K/N Σ_{i<j} s_i²s_j², whereas (3.18) and (4.2) use −K/(2N) Σ_{i,j} s_i²s_j², which differ by a diagonal self-term; this finite-N algebra propagates through Appendices A–B into the final formulas.
  3. [§4.2, Corollary 17 (Eq. (4.74))] For the Katayama-Horiguchi limit, the 1-RSB pressure is printed with +βD in the s² coefficient, while the RS version (4.25) and the annealed result (3.17) have −βD; the equation also uses q̄ instead of q̄₂ in the fluctuation term. This is not a sign-convention ambiguity and should be corrected before the claimed agreement with [35] can be checked.
  4. [Assumption 1 (Eq. (4.6)) and Remark 3] The central derivation relies on unproved concentration of q_ab, q_aa, and m on their means, and Remark 3 concedes that even the existence of the infinite-volume quenched pressure is assumed because Guerra-Toninelli does not apply straightforwardly. Without (4.6), the step from Lemma 1 to Lemma 2 does not close. This is acceptable if the paper is presented as heuristic or conditional, but the abstract and Section 5 state that the analysis is rigorous and 'prove[s] full agreement'; the wording should be softened accordingly.
minor comments (4)
  1. [§4.1, proof of Theorem 1] The sentence 'this procedure returns Eqs. (4.26)–(4.28)' should refer to Eqs. (4.8)–(4.10); Eqs. (4.26)–(4.28) belong to Corollary 11.
  2. [Corollary 22, Eq. (4.105)] The AT-line formula for the BEGC model contains a nested expectation and a ξ defined with −4βKQ, while Theorem 3's tilde-W uses +βKQ; please clarify the convention and check the algebra.
  3. [Various displayed equations] Several displayed equations have typographical issues: (4.82) has an unbalanced parenthesis around the sinh term, and (4.92) uses a fraction whose numerator and denominator are not cleanly separated. These make verification harder.
  4. [Appendix D, Eq. (4.95)] Equation (4.95) is introduced as the value of the derivative K, but the displayed expression contains a dangling logarithmic term and no statement of its regime of validity; please expand the derivation.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular reduction: Guerra interpolation derives the RS/1-RSB pressures from the Hamiltonian, with known replica results recovered only as comparisons; the only self-citation is a non-load-bearing methodological pointer.

full rationale

The derivation chain is self-contained. Theorem 1 is obtained from the interpolating partition function (4.2), with the constants psi, A, B chosen in (4.39) to make the t-derivative (4.33) independent of t; the RS pressure (4.7) then follows from the Fundamental Theorem of Calculus, and the order-parameter equations are stated as the stationarity conditions of that pressure. Existing GS/KH/BEG results are cited after the fact as checks, not used as inputs: the paper explicitly says it 'recovers' them (Corollaries 8-12) by taking limits of its own Theorem 1. The AT line is derived by expanding the paper's own 1-RSB expression (4.54) around theta=1 in Section 4.3 and Appendix D; the citation to the authors' [5] is only a pointer to the expansion technique and is not load-bearing because the expansion is carried out in the manuscript. The stated limitations, Remark 3 and Assumptions 1-2, concede that self-averaging and existence of the thermodynamic limit are assumed rather than proved; that is a correctness risk, not a circularity. The sign disagreement between the KQ term in (4.7) and in (4.8)-(4.10)/(4.30)-(4.31) is an internal algebraic issue to be fixed, but it is not an instance of a prediction being equivalent to its inputs.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The central claim rests on the assumed concentration of order parameters and the assumed existence of the quenched limit. The model parameters J0, J, D, K, h, S, beta are inputs from the model definition. No new physical entities are introduced.

free parameters (1)
  • theta (1-RSB Parisi parameter)
    Introduced in Assumption 2 and Theorem 2 as the weight of the two-peak overlap distribution; no extremization condition over theta is provided, so the 1-RSB free energy is not fully specified.
assumptions (3)
  • domain assumption Existence of the infinite-volume limit of the quenched free energy for Hamiltonian (2.1)
    Remark 3 admits the Guerra-Toninelli approach does not apply straightforwardly and the limit is assumed.
  • domain assumption RS self-averaging of order parameters (Assumption 1)
    Section 4.1 Eq. (4.6): variances of q_ab, q_aa, and m vanish in the thermodynamic limit. Not proven.
  • domain assumption 1-RSB overlap distribution with two peaks (Assumption 2)
    Section 4.2 Eqs. (4.52)-(4.53): P(q_ab) = theta delta(q_ab - qbar_1) + (1-theta) delta(q_ab - qbar_2). Not derived.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Guerra interpolation for inverse freezing." pith.science (2026). https://pith.science/paper/5XWWVS2Q

@misc{pith2026250506202,
  author       = {Pith},
  title        = {Pith review of: Guerra interpolation for inverse freezing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5XWWVS2Q}},
  note         = {Machine review of arXiv:2505.06202}
}
read the original abstract

In these short notes, we adapt and systematically apply Guerra's interpolation techniques on a class of disordered mean-field spin glasses equipped with crystal fields and multi-value spin variables. These models undergo the phenomenon of inverse melting or inverse freezing. In particular, we focus on the Ghatak-Sherrington model, its extension provided by Katayama and Horiguchi, and the disordered Blume-Emery-Griffiths-Capel model in the mean-field limit, deepened by Crisanti and Leuzzi and by Schupper and Shnerb. Once shown how all these models can be retrieved as particular limits of a unique broader Hamiltonian, we study their free energies. We provide explicit expressions of their annealed and quenched expectations, inspecting the cases of replica symmetry and (first-step) broken replica symmetry. We recover several results previously obtained via heuristic approaches (mainly the replica trick) to prove full agreement with the existing literature. As a sideline, we also inspect the onset of replica symmetry breaking by providing analytically the expression of the de Almeida-Thouless instability line for a replica symmetric description: in this general setting, the latter is new also from a physical viewpoint.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

60 extracted references · 59 canonical work pages

  1. [35]

    Katayama and T

    K. Katayama and T. Horiguchi. Ghatak-Sherrington mode l with spin S. Journal of the Physical Society of Japan , 68(12):3901–3910, 1999

    work page 1999

  2. [1]

    Agliari, L

    E. Agliari, L. Albanese, A. Barra, and G. Ottaviani. Repl ica symmetry breaking in neural networks: A few steps toward rigorous results. Journal of Physics A: Mathematical and Theoretical , 53, 2020

    work page 2020

  3. [2]

    Agliari, F

    E. Agliari, F. Alemanno, A. Barra, and A. Fachechi. Dream ing neural networks: rigorous results. Journal of Statistical Mechanics: Theory and Experiment , 2019(8):083503, 2019

    work page 2019

  4. [3]

    Agliari, A

    E. Agliari, A. Barra, P. Sollich, and L. Zdeborova. Machi ne learning and statistical physics: theory, inspiration, application, 2020

    work page 2020

  5. [4]

    Albanese, F

    L. Albanese, F. Alemanno, A. Alessandrelli, and A. Barra . Replica symmetry breaking in dense hebbian neural networks. Journal of Statistical Physics , 189(2):1–41, 2022

    work page 2022

  6. [5]

    Albanese, A

    L. Albanese, A. Alessandrelli, A. Annibale, and A. Barra . About the de Almeida–Thouless line in neural networks. Physica A: Statistical Mechanics and its Applications , 633:129372, 2024. – 39 –

    work page 2024

  7. [6]

    Alberici and a

    D. Alberici and a. et. Annealing and replica-symmetry in deep boltzmann machines. Journal of Statistical Physics, 180:665–677, 2020

    work page 2020

  8. [7]

    Auffinger and C

    A. Auffinger and C. X. Chen. Thouless–Anderson–Palmer equ ations for the Ghatak–Sherrington mean field spin glass model. Journal of Statistical Physics , 184:1–25, 2021

    work page 2021

Show all 60 references
  1. [8]

    Barra, A

    A. Barra, A. Di Biasio, and F. Guerra. Replica symmetry br eaking in mean-field spin glasses through the Hamilton–Jacobi technique. Journal of Statistical Mechanics: Theory and Experiment , 2010(09):P09006, 2010

    2010

  2. [9]

    Barra, G

    A. Barra, G. Genovese, and F. Guerra. The replica symmetr ic approximation of the analogical neural network. Journal of Statistical Physics , 140:784–796, 2010

    2010

  3. [10]

    Barra, G

    A. Barra, G. Genovese, and F. Guerra. Equilibrium stati stical mechanics of bipartite spin systems. Journal of Physics A , 44(24):245002, 2011

    2011

  4. [11]

    Barra, F

    A. Barra, F. Guerra, and E. Mingione. Interpolating the Sherrington–Kirkpatrick replica trick. Philosophical Magazine , 92(1):78, 2012

    2012

  5. [12]

    Bolle, I

    D. Bolle, I. Perez-Castillo, and G. Shim. Optimal capac ity of the Blume-Emery-Griffiths perceptron. Physical Review E , 67:036113, 2003

    2003

  6. [13]

    Bolle and T

    D. Bolle and T. Verbeiren. Thermodynamics of fully conn ected Blume–Emery–Griffiths neural networks. Journal of Physics A , 36:295, 2003

    2003

  7. [14]

    N. Branco. Blume-Emery-Griffiths model in a random cryst al field. Physical Review B , 60(2):0133, 1999

    1999

  8. [15]

    Carleo and a

    G. Carleo and a. et. Machine learning and the physical sc iences, 2019

    2019

  9. [16]

    Castellani and A

    T. Castellani and A. Cavagna. Spin-glass theory for ped estrians. Journal of Statistical Mechanics: Theory and Experiment , 2005(05):P05012, 2005

    2005

  10. [17]

    W.-K. Chen. On the Almeida-Thouless transition line in the Sherrington-Kirkpatrick model with centered gaussian external field. Electronic Communications in Probability , 26:1–9, 2021

    2021

  11. [18]

    Cirillo and F

    E. Cirillo and F. Nardi. Relaxation height in energy lan dscapes: an application to multiple metastable states. Journal of Statistical Physics , 150:1080, 2013

    2013

  12. [19]

    Cirillo and E

    E. Cirillo and E. Olivieri. Metastability and nucleati on for the Blume-Capel model. different mechanisms of transition. Journal of Statistical Physics , 83:473, 1996

    1996

  13. [20]

    Crisanti and L

    A. Crisanti and L. Leuzzi. Spherical 2 + p spin-glass model: An exactly solvable model for glass to spin-glass transition. Phys. Rev. Lett. , 93:217203, Nov 2004

    2004

  14. [21]

    Crisanti and L

    A. Crisanti and L. Leuzzi. Stable solution of the simple st spin model for inverse freezing. Physical review letters, 95(8):087201, 2005

    2005

  15. [22]

    Crisanti and L

    A. Crisanti and L. Leuzzi. Spherical 2 + p spin-glass model: An analytically solvable model with a glass-to-glass transition. Phys. Rev. B , 73:014412, Jan 2006

    2006

  16. [23]

    Crisanti and L

    A. Crisanti and L. Leuzzi. Thermodynamic properties of a full-replica-symmetry-breaking ising spin glass on lattice gas: The random Blume-Emery-Griffiths-Cape l model. Physical Review B , 70:014409, 2014

    2014

  17. [24]

    F. A. da Costa, C. S. Yokoi, and S. R. Salinas. First-orde r transition in a spin-glass model. Journal of Physics A: Mathematical and General , 27(10):3365, 1994

    1994

  18. [25]

    J. R. de Almeida and D. J. Thouless. Stability of the Sher rington-Kirkpatrick solution of a spin glass model. Journal of Physics A: Mathematical and General , 11(5):983, 1978

    1978

  19. [26]

    De Dominicis and I

    C. De Dominicis and I. Giardina. Random fields and spin glasses: a field theory approach . Cambridge University Press, 2006. – 40 –

    2006

  20. [27]

    Feldmann and R

    H. Feldmann and R. Oppermann. Replica symmetry breakin g solution for the fermionic Ising spin-glass and the Ghatak-Sherrington model. Journal of Physics A , 33:1325, 2000

    2000

  21. [28]

    Ghatak and D

    S. Ghatak and D. Sherrington. Crystal field effects in a ge neral S ising spin glass. Journal of Physics C: Solid State Physics , 10(16):3149, 1977

    1977

  22. [29]

    F. Guerra. About the overlap distribution in mean field s pin glass models. International Journal of Modern Physics B , 10(13n14):1675–1684, 1996

    1996

  23. [30]

    F. Guerra. Broken replica symmetry bounds in the mean fie ld spin glass model. Communications in Mathematical Physics , 233:1–12, 2003

    2003

  24. [31]

    F. Guerra. The replica symmetric region in the Sherring ton-Kirkpatrick mean field spin glass model. the almeida-thouless line. arXiv preprint cond-mat/0604674 , 2006

    2006 arXiv

  25. [32]

    Guerra and a

    F. Guerra and a. et. About a solvable mean field model of a g aussian spin glass. Journal of Physics A , 47(15):155002, 2014

    2014

  26. [33]

    Guerra and a

    F. Guerra and a. et. Mean field bipartite spin models trea ted with mechanical techniques. The European Physical Journal B , 87:1, 2014

    2014

  27. [34]

    Guerra and F

    F. Guerra and F. L. Toninelli. The thermodynamic limit i n mean field spin glass models. Communications in Mathematical Physics , 230:71–79, 2002

    2002

  28. [36]

    Kaufman and M

    M. Kaufman and M. Kanner. Random-field Blume-Capel mode l: Mean-field theory. Physics Review B , 42:2378, 1990

    1990

  29. [37]

    Krotov and J

    D. Krotov and J. J. Hopfield. Dense associative memory fo r pattern recognition. Advances in Neural Information Processing Systems , pages 1180–1188, 2016

    2016

  30. [38]

    LeCun, Y

    Y. LeCun, Y. Bengio, and G. Hinton. Deep learning. Nature, 521:436–444, 2015

    2015

  31. [39]

    L. Leuzzi. Spin-glass model for inverse freezing. Philosophical Magazine , 87(3-5):543–551, 2007

    2007

  32. [40]

    Leuzzi, M

    L. Leuzzi, M. Paoluzzi, and A. Crisanti. Random Blume-C apel model on a cubic lattice: First-order inverse freezing in a three-dimensional spin-glass system . Phys. Rev. B , 83:014107, 2011

    2011

  33. [41]

    Marques and a

    M. Marques and a. et. Possible mechanism for cold denatu ration of proteins at high pressure. Physical Review Letters, 91(13):138103, 2003

    2003

  34. [42]

    Morais, M

    C. Morais, M. Lazo, F. Zimmer, and S. Magalhaes. Inverse freezing in the Ghatak-Sherrington model with a random field. Physical Review E—Statistical, Nonlinear, and Soft Matter Physics, 85(3):031133, 2012

    2012

  35. [43]

    Morais, M

    C. Morais, M. Lazo, F. Zimmer, and S. Magalh˜ aes. Invers e transitions in the Ghatak–Sherrington model with bimodal random fields. Physica A: Statistical Mechanics and its Applications , 392(8):1770–1775, 2013

    2013

  36. [44]

    Morais, F

    C. Morais, F. Zimmer, and S. Magalhaes. Inverse freezin g in the hopfield fermionic ising spin glass with a transverse magnetic field. Physics Letters A , 375:689–697, 2011

    2011

  37. [45]

    J.-C. Mourrat. The Parisi formula is a Hamilton–Jacobi equation in Wasserstein space. Canadian Journal of Mathematics , 74(3):607–629, 2022

    2022

  38. [46]

    M´ ezard, G

    M. M´ ezard, G. Parisi, and M. A. Virasoro. Spin glass theory and beyond: An Introduction to the Replica Method and Its Applications , volume 9. World Scientific Publishing Company, 1987

    1987

  39. [47]

    Panchenko

    D. Panchenko. The Parisi ultrametricity conjecture. Annals of Mathematics , pages 383–393, 2013. – 41 –

    2013

  40. [48]

    Parisi and F

    G. Parisi and F. Zamponi. Mean-field theory of hard spher e glasses and jamming. Reviews of Modern Physics, 82(1):789–845, 2010

    2010

  41. [49]

    K. N. Pham and a. et. Glasses in hard spheres with short-r ange attraction. Physical Review E , 69(1):011503, 2004

    2004

  42. [50]

    Plazanet and a

    M. Plazanet and a. et. Inverse freezing in molecular bin ary mixtures of α -cyclodextrin and 4-methylpyridine. Physical Chemistry Chemical Physics , 12(26):7026–7031, 2010

    2010

  43. [51]

    Schupper and N

    N. Schupper and N. Shnerb. Inverse melting and inverse f reezing: a spin model. Physical Review E , 72:046107, 2005

    2005

  44. [52]

    Schupper and N

    N. Schupper and N. M. Shnerb. Spin model for inverse melt ing and inverse glass transition. Physical Review Letters, 93(3):037202, 2004

    2004

  45. [53]

    Y. Sheng. Fluctuation of free energy for the mean-field G hatak-Sherrington Model. arXiv preprint arXiv:2408.16133, 2024

    2024 arXiv

  46. [54]

    Sherrington and S

    D. Sherrington and S. Kirkpatrick. Solvable model of a s pin-glass. Physical review letters , 35(26):1792, 1975

    1975

  47. [55]

    Talagrand

    M. Talagrand. Rigorous low-temperature results for th e mean field p-spins interaction model. Probability theory and related fields , 117(3):303–360, 2000

    2000

  48. [56]

    Talagrand et al

    M. Talagrand et al. Spin glasses: a challenge for mathematicians: cavity and me an field models , volume 46. Springer Science & Business Media, 2003

    2003

  49. [57]

    Temesv´ ari and I

    T. Temesv´ ari and I. Kondor. Field theory for the Almeid a-Thouless transition. arXiv preprint arXiv:2212.01654, 2022

    2022 arXiv

  50. [58]

    F. L. Toninelli. About the Almeida-Thouless transitio n line in the Sherrington-Kirkpatrick mean-field spin glass model. EPL (Europhysics Letters) , 60(5):764, 2002

    2002

  51. [59]

    T. Yokota. Thouless–Anderson–Palmer equations for th e Ghatak–Sherrington mean field spin glass model. Journal of Physics C , 4:2615, 1992

    1992

  52. [60]

    F. M. Zimmer, I. Berger, and S. Magalhaes. Inverse freez ing in the van Hemmen fermionic Ising spin glass with a transverse magnetic field. Physics Letters A , 376:566–572, 2012. – 42 –

    2012

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.