Evaluating the noise resilience of variational quantum algorithms

We simulate the effects of different types noise in state preparation circuits variational quantum algorithms. first use a eigensolver to find ground Hamiltonian presence noise, and adopt two quality measures addition energy, namely fidelity concurrence. then extend task one constructing, with layered circuit ansatz, set general random target states. determine optimal depth for levels observe that algorithms mitigate by adapting optimised parameters. inclusion redundant parameterised gates makes more resilient noise. For such overparameterised sets parameters can result same final noiseless case, which we denote as parameter degeneracy. Numerically, show this degeneracy be lifted some states being significantly than others. also average deviation from is linear level, long small compared circuit-dependent threshold. In region well described stochastic model. Above threshold, optimisation converge largely physical properties true state, so practical applications it critical ensure are below