arxiv: 2605.10326 · v2 · submitted 2026-05-11 · ❄️ cond-mat.stat-mech · physics.comp-ph

Recognition: no theorem link

Statistical mechanics of the N-queens problem

Hai-Jun Liao, Lei Wang, Zong-Yue Liu

Authors on Pith no claims yet

Pith reviewed 2026-05-13 07:27 UTC · model grok-4.3

classification ❄️ cond-mat.stat-mech physics.comp-ph

keywords N-queens problemstatistical mechanicsMonte Carlo simulationthermodynamic integrationground-state entropySimkin constantlattice gasspecific heat

0 comments

The pith

The Simkin constant γ can be recovered from Monte Carlo simulations of the N-queens problem by integrating the specific heat down to zero temperature.

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

This paper treats the N-queens problem as a lattice gas in which queens repel along shared rows, columns and diagonals. Monte Carlo runs up to N=1024 show that the specific heat per queen converges to a single N-independent curve with a finite peak near T=0.235, so there is no phase transition. Because the high-temperature entropy is known exactly, thermodynamic integration of C_v/T from infinity to zero then yields the ground-state entropy per queen and therefore the constant γ. The simulation value 1.946 matches the combinatorial result 1.944 to within 0.1 percent.

Core claim

By mapping the N-queens problem to a lattice gas with repulsive interactions, the authors find that the specific heat per queen converges to a universal function of temperature. Thermodynamic integration of this function from infinite temperature, where the entropy is known exactly, to T=0 recovers the ground-state entropy per queen and hence the Simkin constant γ directly from the Monte Carlo data.

What carries the argument

Convergence of the specific heat per queen C_v/N to a universal, N-independent function of temperature, which permits thermodynamic integration of C_v/T from T=∞ to T=0.

If this is right

The N-queens lattice gas has no thermodynamic phase transition in the large-N limit.
The high-temperature energy per queen is exactly 5/3.
Ground-state entropy can be obtained from simulation without enumerating solutions.
The same integration route supplies an independent numerical check on the Simkin constant γ.

Where Pith is reading between the lines

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

The same integration technique could be applied to other hard constraint problems whose ground-state entropy is difficult to count directly.
Hybrid use of the transfer-matrix tensor network and Monte Carlo might give both exact small-N benchmarks and large-N estimates.
Finite-size scaling of the specific-heat peak could be used to extrapolate γ to infinite N with controlled error.

Load-bearing premise

The specific heat per queen must approach a smooth universal function of temperature that can be integrated accurately to zero without missing non-analyticities or finite-size biases.

What would settle it

If larger-N simulations produced a divergent peak in C_v or an integrated entropy that deviates from the known combinatorial value of γ by more than the reported uncertainty.

Figures

Figures reproduced from arXiv: 2605.10326 by Hai-Jun Liao, Lei Wang, Zong-Yue Liu.

**Figure 3.** Figure 3: FIG. 3. Energy per queen [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

**Figure 2.** Figure 2: FIG. 2. Convergence diagnostics for the Monte Carlo simu [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 4.** Figure 4: FIG. 4. (a) Specific heat per queen [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗

**Figure 5.** Figure 5: FIG. 5. (a) The 17 nonzero elements of the rank-9 site tensor [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗

**Figure 6.** Figure 6: FIG. 6. Tensor network for exact counting of [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗

read the original abstract

We investigate the $N$-queens problem as a lattice gas -- a model in which $N$ queens are placed on an $N \times N$ chessboard with pairwise repulsive interactions along shared rows, columns, and diagonals -- from the perspective of statistical mechanics. The ground states are exactly the $Q(N)$ solutions of the classical $N$-queens problem, with entropy per queen $s_0 \approx \ln N - \gamma$ ($\gamma \approx 1.944$). This entropy reflects a characteristic constraint hierarchy: each successive geometric constraint -- columns, then diagonals -- reduces the entropy from the free-placement value $\ln N$ by a definite constant. We derive the exact high-temperature energy $E/N \to 5/3$ as $N \to \infty$. Extensive Monte Carlo simulations with $10^8$ sweeps per temperature point for $N = 8$--$1024$ reveal that the specific heat per queen $C_v/N$ converges to a universal function of $T$ as $N \to \infty$. The converged curve features a non-divergent peak $C_v^{\max}/N \approx 1.63$ at $T^* \approx 0.235\,J$, establishing the absence of a thermodynamic phase transition. Combined with the trivially exact high-temperature entropy $S(\infty)/N = (1/N) \ln \binom{N^2}{N}$, the convergence of $C_v/N$ enables a thermodynamic integration of $C_v/T$ from $T = \infty$ to $T = 0$ that recovers the ground-state entropy -- and hence the Simkin constant $\gamma$ -- purely from Monte Carlo data. This provides an independent thermodynamic route to a fundamental combinatorial constant. Thermodynamic integration yields $\gamma_{\rm MC} = 1.946 \pm 0.003$ at $N = 1024$, within $0.1\%$ of the precise combinatorial value $\gamma = 1.94400(1)$. We further present a transfer-matrix-based tensor network formulation that encodes the non-attacking constraints into a rank-9 site tensor with 17 nonzero elements, providing a complementary exact-enumeration route.

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.

Desk Editor's Note private letter to a colleague

The paper shows that Monte Carlo on a lattice-gas model of N-queens plus thermodynamic integration recovers the Simkin constant gamma to 0.1% accuracy from finite-T data alone.

read the letter

The core result is that the N-queens problem can be recast as a lattice gas with row, column, and diagonal repulsions. From there the authors derive the exact high-T energy per queen of 5/3, run large Monte Carlo runs up to N=1024, and show that Cv/N collapses onto a smooth, N-independent curve with a modest peak and no divergence. Thermodynamic integration of that curve from infinite T down to zero then yields gamma_MC = 1.946(3), within 0.1% of the accepted combinatorial value. That route is new in the N-queens literature and avoids feeding the known count into the calculation. The numerical evidence for convergence looks solid at the sizes they reach, and the absence of a thermodynamic transition is convincing from the specific-heat data. The tensor-network sketch is a nice side note but clearly secondary. The main soft spot is that the convergence of Cv/N to a universal function is demonstrated numerically rather than proven, so the integration step still rests on an assumption that finite-size corrections and any hidden non-analyticities remain small enough not to bias the zero-T entropy. The finite-size scaling analysis is also only sketched. Readers who work on numerical entropy calculations or on mapping combinatorial problems to statistical mechanics will get the most out of it. The work is clean enough and the central numerical claim holds up well enough that it deserves a serious referee.

Referee Report

2 major / 2 minor

Summary. The manuscript models the N-queens problem as a lattice gas with repulsive interactions along rows, columns, and diagonals. It derives the exact high-temperature limit E/N → 5/3, performs Monte Carlo simulations (10^8 sweeps) for N = 8 to 1024 showing that Cv/N converges to a universal, N-independent curve with a non-divergent peak at T* ≈ 0.235J, establishes the absence of a thermodynamic phase transition, and uses thermodynamic integration of Cv/T from T = ∞ (with exact high-T entropy) to T = 0 to recover the ground-state entropy and thus the Simkin constant γ_MC = 1.946 ± 0.003, within 0.1% of the combinatorial value 1.94400(1). A rank-9 tensor-network formulation with 17 nonzero elements is also introduced.

Significance. If the reported convergence of Cv/N holds under finite-size extrapolation, the work supplies an independent thermodynamic route to the Simkin constant γ that relies only on Monte Carlo data and the exact high-T limit, rather than direct combinatorial enumeration. The large-scale simulations, exact high-T derivation, and close numerical agreement constitute clear strengths and demonstrate that the model has no thermodynamic transition. This approach may generalize to other constrained combinatorial problems.

major comments (2)

[Thermodynamic integration] § on thermodynamic integration: the claim that finite-N corrections to the integrated entropy remain small at N = 1024 (yielding 0.1% accuracy on γ) rests on visual convergence of Cv/N; a quantitative extrapolation of the integral versus 1/N or 1/N^2, with error bands propagated from the Monte Carlo statistics, is needed to confirm that residual finite-size bias does not exceed the stated ±0.003 uncertainty.
[Tensor network formulation] Tensor-network section: the rank-9 site tensor with 17 nonzero elements is stated to encode the non-attacking constraints, but neither the explicit tensor entries nor the contraction/transfer-matrix algorithm used for enumeration are supplied, preventing independent verification of this complementary exact route.

minor comments (2)

[Figures] Figure 2 (Cv/N curves): the legend should explicitly list the N values and state that T is in units of J; the peak location T* ≈ 0.235J should be marked on the plot.
[High-temperature limit] The high-T entropy formula S(∞)/N = (1/N) ln binom(N^2, N) is used without derivation; a one-line justification that this is the exact non-interacting limit would aid clarity.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading, positive assessment, and constructive suggestions. We address each major comment below and will revise the manuscript accordingly.

read point-by-point responses

Referee: the claim that finite-N corrections to the integrated entropy remain small at N = 1024 (yielding 0.1% accuracy on γ) rests on visual convergence of Cv/N; a quantitative extrapolation of the integral versus 1/N or 1/N^2, with error bands propagated from the Monte Carlo statistics, is needed to confirm that residual finite-size bias does not exceed the stated ±0.003 uncertainty.

Authors: We agree that a quantitative extrapolation would provide stronger confirmation of convergence. In the revised manuscript we will add a finite-size extrapolation of the integrated entropy (equivalently γ) versus 1/N, together with error bands obtained by propagating the Monte Carlo statistical uncertainties through the thermodynamic integration. This will explicitly verify that the residual bias at N=1024 lies within the quoted ±0.003 uncertainty. revision: yes
Referee: the rank-9 site tensor with 17 nonzero elements is stated to encode the non-attacking constraints, but neither the explicit tensor entries nor the contraction/transfer-matrix algorithm used for enumeration are supplied, preventing independent verification of this complementary exact route.

Authors: We acknowledge the omission. In the revised manuscript we will supply the explicit 17 nonzero entries of the rank-9 site tensor (in a compact notation) and a concise description of the transfer-matrix contraction procedure used to obtain the exact enumeration, thereby enabling independent verification. revision: yes

Circularity Check

0 steps flagged

No significant circularity: exact high-T limits plus independent MC integration

full rationale

The derivation begins with two exact inputs: the high-temperature energy limit E/N → 5/3 (derived from pairwise interaction counting) and the exact high-T entropy S(∞)/N = (1/N) ln binom(N²,N). Monte Carlo simulations for N up to 1024 independently sample the specific heat C_v(T) and demonstrate its convergence to an N-independent curve with a finite peak. Thermodynamic integration of C_v/T from ∞ to 0 then produces S(0)/N and the extracted γ_MC, which is compared post hoc to the known combinatorial value. No equation reduces to a fitted parameter renamed as a prediction, no self-citation supplies a load-bearing uniqueness theorem, and the tensor-network formulation is presented as an independent exact-enumeration route. The numerical agreement serves as validation rather than an input, leaving the central claim self-contained against external combinatorial benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests on the convergence of the specific heat to a universal curve and on the validity of thermodynamic integration across the entire temperature range. No new particles or forces are introduced; the model uses standard repulsive pair interactions.

axioms (2)

standard math The high-temperature energy per queen approaches exactly 5/3 as N → ∞
Obtained by enumerating the average number of attacking pairs in the high-T limit.
domain assumption The specific heat per queen converges to an N-independent function for large N
Supported by numerical collapse but not proven analytically.

pith-pipeline@v0.9.0 · 5724 in / 1477 out tokens · 41151 ms · 2026-05-13T07:27:10.851892+00:00 · methodology

Review history (2 revisions) →

discussion (0)

Reference graph

Works this paper leans on

19 extracted references · 19 canonical work pages

[1]

Bezzel, Schachfreund, Berliner Schachzeitung3, 636 (1848)

M. Bezzel, Schachfreund, Berliner Schachzeitung3, 636 (1848)

work page
[2]

Simkin, The number ofn-queens configurations, Adv

M. Simkin, The number ofn-queens configurations, Adv. Math.427, 109127 (2023)

work page 2023
[3]

Bowtell and P

C. Bowtell and P. Keevash, Then-queens problem, arXiv:2109.08083 (2021)

work page arXiv 2021
[4]

Luria and M

Z. Luria and M. Simkin, A lower bound for then-queens problem, arXiv:2105.11431 (2021)

work page arXiv 2021
[5]

Nobel, A

P. Nobel, A. Agrawal, and S. Boyd, Computing tighter bounds on then-queens constant via Newton’s method, Optim. Lett.17, 1229–1240 (2023)

work page 2023
[6]

Yao and Y

G. Yao and Y. Li, High-performanceN-queens solver on GPU: iterative DFS with zero bank conflicts, 9 arXiv:2511.12009 (2025)

work page arXiv 2025
[7]

D. E. Knuth,The Art of Computer Programming, Vol. 4B (Addison-Wesley, 2022), Sec. 7.2.2

work page 2022
[8]

Zhang and J

C. Zhang and J. Ma, Counting solutions for theN-queens and Latin-square problems by efficient Monte Carlo sim- ulations, Phys. Rev. E79, 016703 (2009)

work page 2009
[9]

Polson and V

N. Polson and V. Sokolov, CountingNqueens, arXiv:2407.08830 (2024)

work page arXiv 2024
[10]

Kawasaki, Diffusion constants near the critical point for time-dependent Ising models

K. Kawasaki, Diffusion constants near the critical point for time-dependent Ising models. I, Phys. Rev.145, 224 (1966)

work page 1966
[11]

Metropolis, A

N. Metropolis, A. W. Rosenbluth, M. N. Rosenbluth, A. H. Teller, and E. Teller, Equation of state calcula- tions by fast computing machines, J. Chem. Phys.21, 1087 (1953)

work page 1953
[12]

Efron,The Jackknife, the Bootstrap, and Other Re- sampling Plans(SIAM, Philadelphia, 1982)

B. Efron,The Jackknife, the Bootstrap, and Other Re- sampling Plans(SIAM, Philadelphia, 1982)

work page 1982
[13]

Binder and D

K. Binder and D. W. Heermann,Monte Carlo Simulation in Statistical Physics(Springer, Berlin, 5th ed., 2010)

work page 2010
[14]

Kourtis, C

N. Kourtis, C. Chamon, E. R. Mucciolo, and A. E. Ruck- enstein, Fast counting with tensor networks, SciPost Phys.7, 060 (2019)

work page 2019
[15]

Xiang,Density Matrix and Tensor Network Renormal- ization(Cambridge University Press, Cambridge, 2024)

T. Xiang,Density Matrix and Tensor Network Renormal- ization(Cambridge University Press, Cambridge, 2024)

work page 2024
[16]

Fr¨ owis, V

F. Fr¨ owis, V. Nebendahl, and W. D¨ ur, Tensor operators: Constructions and applications for long-range interaction systems, Phys. Rev. A81, 062337 (2010)

work page 2010
[17]

Nishino and K

T. Nishino and K. Okunishi, Corner transfer matrix renormalization group method, J. Phys. Soc. Jpn.65, 891–894 (1996)

work page 1996
[18]

Vanderstraeten, B

L. Vanderstraeten, B. Vanhecke, and F. Verstraete, Residual entropies for three-dimensional frustrated spin systems with tensor networks, Phys. Rev. E98, 042145 (2018)

work page 2018
[19]

Liu, Simulation code and data for theN- queens lattice gas,https://github.com/LiuZY613/ nqueen-lattice-gas

Z.-Y. Liu, Simulation code and data for theN- queens lattice gas,https://github.com/LiuZY613/ nqueen-lattice-gas

work page