Moduli Spaces of Gauge Theories from Dimer Models: Proof of the Correspondence
Sebastian Franco, David Vegh
TL;DR
This work resolves the Mathematical Dimer Conjecture for toric quivers embedded on a two-torus by showing a one-to-one correspondence between GLSM fields and perfect matchings in the brane tiling, and by proving that the toric diagram equals the Newton polygon of the dimer's characteristic polynomial $P(x,y)$. The authors reformulate F-term equations using gauge transformations and magnetic flux coordinates, yielding a natural identification of GLSM variables with perfect matchings and deriving toric coordinates from height-function slopes. They connect the Forward Algorithm's GLSM data to the dimer framework via the KT product and a left-inverse relation $T=K_L^{-1}K$, demonstrating that height changes $(h_x,h_y)$ reproduce the toric diagram coordinates and thus encode the moduli space geometry. This result strengthens dimers as a robust, computationally efficient tool for analyzing toric quiver gauge theories and their AdS/CFT duals, with clear paths for extending to non-toric phases and higher-dimensional generalizations.
Abstract
Recently, a new way of deriving the moduli space of quiver gauge theories that arise on the world-volume of D3-branes probing singular toric Calabi-Yau cones was conjectured. According to the proposal, the gauge group, matter content and tree-level superpotential of the gauge theory is encoded in a periodic tiling, the dimer graph. The conjecture provides a simple procedure for determining the moduli space of the gauge theory in terms of perfect matchings. For gauge theories described by periodic quivers that can be embedded on a two-dimensional torus, we prove the equivalence between the determination of the toric moduli space with a gauged linear sigma model and the computation of the Newton polygon of the characteristic polynomial of the dimer model. We show that perfect matchings are in one-to-one correspondence with fields in the linear sigma model. Furthermore, we prove that the position in the toric diagram of every sigma model field is given by the slope of the height function of the corresponding perfect matching.
