Exactly-solvable models as benchmarks for VQE

Perhaps the most promising application of Noisy Intermediate Scale Quantum (NISQ) computers is the Variational Quantum Eigensolver (VQE). Due to their namesake noise, NISQ computers performing VQE will need benchmarks to interpret their results. Exactly solvable models such as the Lipkin-Meshkov-Glick (LMG) model, a simple nuclear model of N fermions, can provide such benchmarks. We give circuits that produce low-N LMG eigenstates on a quantum computer with gate and qubit costs suited to the NISQ era. Further, we discuss how we might generalize the circuits for simulations of a higher number of particles.