Free-energy asymptotic expansion conjecture for the square-lattice Potts eigenvalues

Let qq be the Potts-model parameter, let mm be a positive integer, and let λ,(m)\lambda_{\star,\neq}(m) and λ,=(m)\lambda_{\star,=}(m) denote the relevant dominant eigenvalues for the \neq and == boundary conditions. There exist polynomials C1,C2,C_1,C_2,\ldots with rational coefficients, each of degree 11, such that

logλ,(m)qm=k=1mCk(m)qk+O(q(m+1)),\log \frac{\lambda_{\star,\neq}(m)}{q^m}=\sum_{k=1}^m C_k(m)q^{-k}+O(q^{-(m+1)}), logλ,=(m)qm=k=1mCk(m)qk+(1)m+1qm+O(q(m+1)).\log \frac{\lambda_{\star,=}(m)}{q^m}=\sum_{k=1}^m C_k(m)q^{-k}+(-1)^{m+1}q^{-m}+O(q^{-(m+1)}).

Free-energy expansion conjecture. The two displayed asymptotic expansions hold. This has been verified for m10m\le 10 using the explicitly computed polynomials C1,,C10C_1,\ldots,C_{10}. The conjecture refines the proved leading-order statement for the difference of the two eigenvalues, but remains unproved in general.

Sources & referencesView supporting material

Primary source

Jesús Salas and Alan D. Sokal, “Transfer matrices and partition-function zeros for antiferromagnetic Potts models. VI. Square lattice with special boundary conditions”, arXiv:1002.3761 (2011).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.