The Order of Free Energy Fluctuations in the Critical Sherrington-Kirkpatrick Model Revisited
Pith reviewed 2026-06-26 13:34 UTC · model grok-4.3
The pith
The variance of the free energy at the critical point of the Sherrington-Kirkpatrick model is at most (1/4) log N plus constant and at least (1/2) log log log N minus constant.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
We prove that Var(F_N(1)) ≤ (1/4) log N + C with C independent of N, and that Var(F_N(1)) ≥ (1/2) log log log N - C.
What carries the argument
The free energy F_N(1) given by the logarithm of the partition function whose couplings are i.i.d. Gaussians, evaluated at the critical inverse temperature β=1.
If this is right
- Fluctuations remain of logarithmic order rather than staying bounded or growing as a power of N.
- The upper bound confirms the scaling order conjectured in the physics literature for the critical regime.
- The variance diverges, so the free energy is not self-averaging at criticality.
- The model with standard Gaussian disorder satisfies these explicit bounds.
Where Pith is reading between the lines
- If the upper bound is asymptotically sharp, the variance must grow like c log N for some c at most 1/4.
- The extremely slow lower bound implies that any numerical confirmation of divergence would require astronomically large N.
- The same variance bounds may constrain the possible limiting processes for the free-energy field in related mean-field models.
- Techniques used here could extend to other critical temperatures or to models with non-Gaussian disorder.
Load-bearing premise
That β=1 marks the critical inverse temperature separating the paramagnetic and spin-glass phases in the Sherrington-Kirkpatrick model.
What would settle it
Numerical computation of Var(F_N(1)) for a sequence of N showing either that the variance exceeds (1/4) log N + any fixed C or that it stays bounded as N grows.
read the original abstract
We study the fluctuations of the free energy of the Sherrington-Kirkpatrick model at the critical inverse temperature $\beta=1$. We prove that \[ \operatorname{Var}(F_N(1))\leq \frac14\log N+C, \] with $C$ independent of $N$. This gives a logarithmic upper bound, in agreement with the order predicted in the physics literature. We also prove that the critical variance diverges: more precisely, \[ \operatorname{Var}(F_N(1)) \geq \frac12\log\log\log N - C . \]
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript studies fluctuations of the free energy F_N(1) in the Sherrington-Kirkpatrick model at the critical inverse temperature β=1. It claims to prove the upper bound Var(F_N(1)) ≤ (1/4) log N + C (C independent of N) and the lower bound Var(F_N(1)) ≥ (1/2) log log log N - C, establishing both a logarithmic upper bound matching physics predictions and divergence of the variance.
Significance. If the proofs are correct, the results rigorously confirm the logarithmic order of free energy fluctuations at criticality with explicit constants, providing the first mathematical verification of the physics literature's predicted scaling together with a quantitative lower bound on the divergence rate. This strengthens the mathematical foundation for critical behavior in mean-field spin glasses.
Simulated Author's Rebuttal
We thank the referee for their careful reading and for the positive assessment of the significance of our results. The report provides a concise summary of our claims (O(log N) upper bound and Omega(log log log N) lower bound on Var(F_N(1)) at criticality) but lists no specific major comments. The 'uncertain' recommendation appears to reflect a general need to verify the proofs rather than any identified issues. We are happy to supply additional details or clarifications on any aspect of the arguments if requested.
Circularity Check
No significant circularity detected
full rationale
The paper states direct mathematical proofs of explicit bounds Var(F_N(1)) ≤ (1/4) log N + C and Var(F_N(1)) ≥ (1/2) log log log N - C for the standard SK model at the accepted critical value β=1. No fitted parameters, self-definitional constructions, ansatzes, or load-bearing self-citations appear in the abstract or stated claims. The derivation is a self-contained sequence of probabilistic estimates on the log-partition function with i.i.d. Gaussian couplings; the results do not reduce to their own inputs by construction.
Axiom & Free-Parameter Ledger
axioms (1)
- domain assumption Standard definition of the SK Hamiltonian with i.i.d. Gaussian couplings and the critical inverse temperature beta=1.
Reference graph
Works this paper leans on
-
[1]
Aizenman, J
M. Aizenman, J. L. Lebowitz and D. Ruelle. Some rigorous results on the Sherrington-Kirkpatrick spin glass model.Comm. Math. Phys.112, 3–20 (1987)
1987
-
[2]
G. W. Anderson, A. Guionnet and O. Zeitouni. An Introduction to Random Matri- ces.Cambridge Studies in Advanced Mathematics118, Cambridge University Press, Cambridge (2010)
2010
-
[3]
Aspelmeier
T. Aspelmeier. Free-Energy Fluctuations and Chaos in the Sherrington-Kirkpatrick Model.Phys. Rev. Lett.100, 117205 (2008)
2008
-
[4]
Auffinger and W.-K
A. Auffinger and W.-K. Chen. The Parisi formula has a unique minimizer.Comm. Math. Phys.335, 1429–1444 (2015)
2015
-
[5]
Baik and J
J. Baik and J. O. Lee. Fluctuations of the free energy of the spherical Sherrington– Kirkpatrick model,J. Stat. Phys.165, no. 2, 185–224 (2016)
2016
-
[6]
Brennecke, A
C. Brennecke, A. Schertzer, C. Xu and H.-T. Yau. The Two Point Function of the SK Model without External Field at High Temperature.Prob. Math. Phys.5, no. 1, 131–175 (2024)
2024
-
[7]
Brennecke, A
C. Brennecke, A. Schertzer and C. Van Dam. Spin covariance fluctuations in the SK model at high temperature.Electron. Commun. Probab.30, Paper No. 9, 14 (2025)
2025
-
[8]
S. Chatterjee. Disorder chaos and multiple valleys in spin glasses.arXiv:0907.3381
-
[9]
W.-K. Chen. A Gaussian convexity for logarithmic moment generating functions with applications in spin glasses,Ann. Inst. Henri Poincar´ e Probab. Stat.62, no. 1, 195–206 (2026)
2026
-
[10]
Chen and W.-K
W.-K. Chen and W.-K. Lam. Order of Fluctuations of the Free Energy in the SK Model at Critical Temperature.ALEA, Lat. Am. J. Probab. Math. Stat.16, 809–816 (2019)
2019
-
[11]
P. S. Dey and T. Kang. Fluctuations for the Sherrington-Kirkpatrick spin glass model near the critical temperature.Preprint, arXiv:2603.05636 (2026)
arXiv 2026
-
[12]
F. Guerra. Broken Replica Symmetry Bounds in the Mean Field Spin Glass Model. Comm. Math. Phys.233, 1–12 (2003)
2003
-
[13]
Guerra and F
F. Guerra and F. L. Toninelli. The thermodynamic limit in mean field spin glass modelsCommunications in Mathematical Physics230, 71–79 (2002)
2002
-
[14]
B. Landon. Free energy fluctuations of the two-spin spherical SK model at critical temperature,J. Math. Phys.63, no. 3, Paper No. 033301, 22 (2022). 17
2022
-
[15]
Panchenko
D. Panchenko. The Sherrington-Kirkpatrick Model.Springer Monographs in Math- ematics(2013)
2013
-
[16]
G. Parisi. Infinite number of order parameters for spin-glasses.Phys. Rev. Lett.43, 1754–1756 (1979)
1979
-
[17]
G. Parisi. A sequence of approximate solutions to the SK model for spin glasses.J. Phys. A13, L-115 (1980)
1980
-
[18]
Parisi and T
G. Parisi and T. Rizzo. Phase diagram and large deviations in the free-energy of mean-field spin-glasses.Phys. Rev. B79, 134205 (2009)
2009
-
[19]
Sherrington and S
D. Sherrington and S. Kirkpatrick. Solvable model of a spin glass.Phys. Rev. Lett. 35, 1792–1796 (1975)
1975
-
[20]
Talagrand
M. Talagrand. The Parisi formula.Ann. of Math.163, no. 1, 221–263 (2006)
2006
-
[21]
Talagrand
M. Talagrand. Mean Field Models for Spin Glasses. Volume I: Basic Examples. A Series of Modern Surveys in Mathematics54,Springer Berlin, Heidelberg(2010)
2010
-
[22]
Talagrand
M. Talagrand. Mean Field Models for Spin Glasses. Volume II: Advanced Replica- Symmetry and Low Temperature. A Series of Modern Surveys in Mathematics54, Springer Verlag Berlin-Heidelberg(2011). 18
2011
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.