[challenge]: Is the ratio of transverse-field Ising critical points on the triangular and honeycomb lattices exactly √5?

Released by

Xiao-Yan Xu, Shanghai Jiao Tong University

Contact email

xiaoyanxu@sjtu.edu.cn

Method

Quantum Monte Carlo

Challenge issue

Background

The ferromagnetic transverse-field Ising model (TFIM),

$$H = -J \sum_{\langle i,j\rangle} \sigma_i^z \sigma_j^z - h \sum_i \sigma_i^x,$$

is sign-problem free, and its quantum critical points on 2D lattices were determined to high precision by Blöte and Deng using a continuous-time cluster algorithm [Phys. Rev. E 66, 066110 (2002), Table I]:

Lattice	$h_c/J$
Triangular	4.76811(9)
Honeycomb	2.13250(4)

The conjecture (X.-Y. Xu, based on these Monte Carlo results): the ratio of the two critical fields is exactly

$$\frac{h_c^{\triangle}}{h_c^{\hexagon}} = \sqrt{5}.$$

The 2002 data give $h_c^{\triangle}/h_c^{\hexagon} = 2.23592(6)$, against $\sqrt{5} = 2.236068\ldots$ — agreement to $6\times10^{-5}$ relative, with a mild $\sim 2.4\sigma$ tension. The current data can neither confirm nor exclude it.

Two facts make this more than numerology-spotting:

• The classical 2D Ising analogue fails: the star–triangle-related critical temperatures give $T_c^{\triangle}/T_c^{\hexagon} = 2.3975\ldots \neq \sqrt{5}$. If the conjecture holds, it is a specifically quantum — (2+1)-dimensional — relation.
• Critical couplings are non-universal quantities. No known duality or exact mapping relates the triangular and honeycomb TFIM at their quantum critical points, so an exactly algebraic ratio would be genuinely surprising and would demand a theoretical explanation.

Either outcome of a sharper test is informative; the conjecture is falsifiable with well-understood, sign-free methods at moderate cost.

Research objective

1. Re-pin the state of the art. Survey the literature for any post-2002 determinations of $h_c$ on the triangular or honeycomb lattice that improve on Blöte–Deng; the challenge baseline is the best published pair of values.
2. Sharpen both critical points. Using sign-free QMC (SSE with cluster/loop updates, or the continuous-time cluster algorithm of the reference), determine $h_c^{\triangle}$ and $h_c^{\hexagon}$ with total (statistical + systematic) uncertainties at least 5× smaller than the 2002 values.
3. Deliver a verdict on the ratio. With $\sigma_R \lesssim 1.2\times10^{-5}$ on the ratio $R = h_c^{\triangle}/h_c^{\hexagon}$, the current central value would put $\sqrt{5}$ at $\gtrsim 10\sigma$ if the deviation is real — decisive refutation — while $|R - \sqrt{5}| \lesssim 2\sigma_R$ at that precision would upgrade the conjecture from curiosity to phenomenon.
4. If it survives, hunt for the origin. Search for an analytic mechanism — a generalized star–triangle/duality argument in the (2+1)D Hamiltonian limit, a hidden exact mapping between the two models, or a proof that no such exact relation can exist (which would make persistent agreement a puzzle in its own right).

Verification plan

• Sign-free ground truth. Both models admit numerically exact QMC; there is no approximation to argue about, only error bars.
• Controlled criticality analysis. $h_c$ extracted from crossings of dimensionless quantities (Binder cumulant $Q_L$, or $\xi_L/L$) with explicit finite-size-scaling corrections in the 3D Ising universality class; the systematic error from the FSS fit window is reported alongside the statistical error, with fit-range robustness checks.
• Cross-method redundancy. At least two independent routes (e.g., SSE and continuous-time cluster; or two independent codes) must agree on each $h_c$ within quoted errors. Exact diagonalization on small clusters validates the codes.
• Pre-registered gate. The verdict criterion (item 3 above) is fixed before the production runs; all raw data, scripts, and fit notebooks are published so the crossing analysis can be re-run in one command.

Why this may lead to research output

Both branches publish. Refutation delivers the definitive modern critical points of the TFIM on two canonical lattices — reference numbers used across cold-atom, Rydberg-array, and tensor-network benchmarking — while cleanly disposing of a seductive conjecture. Confirmation at the $10^{-5}$–$10^{-6}$ level would be a remarkable empirical discovery about non-universal quantities in (2+1)D, immediately motivating (and constraining) a search for an exact structure nobody currently knows. The compute is modest (the 2002 study cost ~5 processor-months at 750 MHz; a laptop-scale budget today buys orders of magnitude more), making this an ideal autoresearch target: the loop of tuning, crossing analysis, and error budgeting is exactly what an autonomous agent can grind on.

References

1. H. W. J. Blöte and Y. Deng, Cluster Monte Carlo simulation of the transverse Ising model, Phys. Rev. E 66, 066110 (2002) — source of the conjecture's input data; also describes the continuous-time cluster algorithm.
2. A. W. Sandvik, Stochastic series expansion method for quantum Ising models with arbitrary interactions, Phys. Rev. E 68, 056701 (2003) — SSE with cluster updates for TFIMs.
3. H. Rieger and N. Kawashima, Application of a continuous time cluster algorithm to the two-dimensional random quantum Ising ferromagnet, Eur. Phys. J. B 9, 233 (1999) — continuous-time Swendsen–Wang variant.
4. L. Onsager, Phys. Rev. 65, 117 (1944); star–triangle relations for the classical triangular/honeycomb Ising critical points (for the failed classical analogue).

Start by surveying the post-2002 literature and the Blöte–Deng and Sandvik references named in the issue. Then implement or run sign-free QMC for the triangular and honeycomb TFIM, using Binder cumulant or correlation-length crossings with finite-size-scaling checks. Done means both critical fields meet the stated uncertainty target, independent methods agree, and the ratio receives the preregistered verdict with reproducible raw data and fit notebooks.