A Gauge Sign Rule for Quantum Rotor Networks

Abstract

The sign problem is the non-positive path-integral weight that makes a quantum system hard to simulate classically. It shows up in problems that look unrelated, from fermion antisymmetry and geometric frustration to real-time evolution, topological angles, and finite density. We study it here for compact continuous variables, modeled as networks of quantum rotors, $\hat H=\sum_i 4E_C(\hat n_i-n_{g,i})^2-\sum_{\langle ij\rangle} E_{ij}\cos(\hatϕ_i-\hatϕ_j-θ_{ij})$. For these, a single gauge-invariant quantity controls the obstruction, the frustration flux through the independent loops of the interaction graph. We prove a sign rule, the rotor analogue of Marshall's. The charge-basis Hamiltonian is sign-free if and only if every loop flux vanishes modulo $2π$. Exact diagonalization and density-matrix renormalization then show that the sign cost is a gauge-invariant function of the flux. For a single loop it vanishes at zero flux, peaks at $π$, and is exponentially small in the loop's perimeter. Summed over many loops it is extensive, growing with system size. The same loop flux generates the sign of three problems usually treated apart, namely the Mott transition (through the slave-rotor mapping), compact $U(1)$ lattice gauge theory, and frustrated continuous optimization. A superconducting rotor array realizes both the model and its sign natively.