发表机构
Lake Zurich High School; Virginia Tech（莱克齐尔高中; 弗吉尼亚理工大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究以线性H3分子为例，考察噪声对ADAPT-VQE算子选择步骤的影响，并证明结合动态解耦、零噪声外推和泡利扭曲等误差缓解技术可恢复收敛，提升算法在近期量子硬件上的性能。
AI 中文摘要
硬件噪声已被证明会显著影响ADAPT-VQE（一种基态制备算法）的精度。虽然先前的工作研究了噪声对其参数优化步骤的影响，但其对关键的算子选择步骤的影响相对而言仍未得到探索。在本工作中，我们以线性H$_3$分子作为测试案例，考察了多种噪声通道对该步骤的影响。我们表明，尽管选择准则对某些噪声具有天然的鲁棒性，但相干和非相干噪声在足够高的噪声率下都能阻止收敛。我们采用了量子误差缓解技术——动态解耦、零噪声外推和泡利扭曲——并表明当这些技术适当组合时，能够恢复成功的收敛曲线。我们的结果凸显了误差缓解如何能提升ADAPT-VQE的性能并在硬件噪声存在下实现收敛，为在近期量子硬件上实现该算法提供了宝贵的见解。
英文摘要
Hardware noise has been shown to significantly impact the accuracy of ADAPT-VQE, a ground state preparation algorithm. While previous work has studied the impact of noise on its parameter optimization step, its impact on the critical operator selection step remains comparatively unexplored. In this work, we examine the impact of a variety of noise channels on this step, using a linear H$_3$ molecule as a test case. We show that, despite the selection criterion's natural resilience to some noise, both coherent and incoherent noise can prevent convergence for sufficiently high noise rates. We employ quantum error mitigation techniques--dynamical decoupling, zero noise extrapolation, and Pauli twirling--and show that when combined appropriately, these techniques are capable of restoring a successful convergence profile. Our results highlight how error mitigation can improve the performance of ADAPT-VQE and enable convergence in the presence of hardware noise, offering valuable insights into the implementation of the algorithm on near-term quantum hardware.
