We describe a resource-efficient approach to studying many-body quantum states on noisy, intermediate-scale quantum devices. We employ a sequential generation model that allows us to bound the range of correlations in the resulting many-body quantum states. From this, we characterize situations where the estimation of local observables does not require the preparation of the entire state. Instead smaller patches of the state can be generated from which the observables can be estimated. This can potentially reduce circuit size and number of qubits for the computation of physical properties of the states. Moreover, we show that the effect of noise decreases along the computation. Our results apply to a broad class of widely studied tensor network states and can be directly applied to near-term implementations of variational quantum algorithms.

我们描述了一种资源高效的方法来研究嘈杂的中级量子器件上的多体量子态。我们采用了顺序生成模型，该模型使我们能够在生成的多体量子态中限制相关范围。据此，我们描述了不需要局部状态的估计就可以估计局部可观测量的情况。取而代之的是，可以生成较小的状态补丁，从中可以估计可观察到的状态。这可以潜在地减小电路尺寸和量子位数量，以用于计算状态的物理特性。此外，我们表明，噪声的影响随着计算而降低。我们的结果适用于广泛研究的张量网络状态的广泛类别，并且可以直接应用于变分量子算法的近期实现。