Positivity and periodicity of \(Q\)-systems in the WZW fusion ring (Q2400520)

This interesting paper is motivated by conjectures of \textit{A. N. Kirillov} [J. Sov. Math. 47, No. 2, 2450--2459 (1987; Zbl 0685.33014)] that the characters of certain finite-dimensional modules for a simple Lie algebra \(\mathfrak{g}\) vanish when evaluated at certain weights of \(\mathfrak{g}\) and are positive real numbers when evaluated at other specified weights. The paper concludes with a proof of these conjectures, as well as of more general conjectures by Kuniba, Nakanishi, and Suzuki [\textit{A. Kuniba} et al., J. Phys. A, Math. Theor. 44, No. 10, Article ID 103001, 146 p. (2011; Zbl 1222.82041)], in the case that \(\mathfrak{g}\) is of classical type. It turns out that the numerical values which Kirillov conjectured to be zero or positive are the quantum dimensions of certain elements in the fusion rings \(\mathrm{Fus}_k(\mathfrak{g})\) associated to the modular tensor categories of representations of the affine Lie algebra \(\widehat{\mathfrak{g}}\) at positive integer levels \(k\). Specifically, these elements of \(\mathrm{Fus}_k(\mathfrak{g})\) are images of Kirillov-Reshetikhin modules for the quantum affine algebra \(U_q(\widehat{\mathfrak{g}})\) under the homomorphisms of fusion rings \[ \begin{tikzcd} \operatorname{Rep}U_q(\widehat{\mathfrak{g}}) \ar[r,"\mathrm{res}"] & \operatorname{Rep}U_q(\mathfrak{g}) \ar[r,"\beta_k"] & \operatorname{Fus}_k(\mathfrak{g}) \end{tikzcd}, \] where \(\mathrm{res}\) indicates restriction from a \(U_q(\widehat{\mathfrak{g}})\)-module to a \(U_q(\mathfrak{g})\)-module and \(\beta_k\) is a surjective ring homomorphism defined using the action of the affine Weyl group on the level-\(k\) affinizations of dominant integral weights of \(\mathfrak{g}\). Thus the conjectures of Kirillov and Kuniba-Nakanishi-Suzuki can be approached by studying the images of Kirillov-Reshetikhin modules in \(\mathrm{Fus}_k(\mathfrak{g})\). The centerpiece of the paper is Conjecture 3.10 which largely lifts the conjectures of Kirillov and Kuniba-Nakanishi-Suzuki to \(\mathrm{Fus}_k(\mathfrak{g})\). Essentially, the conjecture is that the images in \(\mathrm{Fus}_k(\mathfrak{g})\) of certain Kirillov-Reshetikhin modules are either \(0\) or positive elements, and that these determine the images of other Kirillov-Reshetikhin modules via a periodicity relation involving the affine Weyl group. The author proves these conjectures for some cases. He also proves the original conjectures of Kirillov and Kuniba-Nakanishi-Suzuki for \(\mathfrak{g}\) of classical type using these partial results together with solutions to \(Q\)-systems, which are non-linear recursive relationships satisfied by the characters of Kirillov-Reshetikhin modules, and thus also by the images of Kirillov-Reshetikhin modules in \(\mathrm{Fus}_k(\mathfrak{g})\) and their quantum dimensions. The key point is that the author is able to construct positive solutions of \(Q\)-systems in \(\mathrm{Fus}_k(\mathfrak{g})\), which then induce positive real solutions of \(Q\)-systems in \(\mathbb{C}\) that can be shown to equal the quantum dimensions used by Kirillov.