arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2604.24580quant-ph
收藏

谱隙感知斜坡QAOA

Spectral Gap Informed Ramp QAOA

National Quantum Computing Centre, Rutherford Appleton Laboratory(国家量子计算中心,拉瑟福德·阿普尔顿实验室);Quantum Software Lab, School of Informatics, The University of Edinburgh(爱丁堡大学信息学院量子软件实验室)

Kieran McDowall, Konstantinos Georgopoulos, Petros Wallden

AI总结:

提出谱隙感知斜坡QAOA(SGIR-QAOA),通过利用绝热哈密顿量的谱隙信息构造平滑参数调度,在恒定深度下提升Grover问题性能并缩短达到最优解概率所需深度,且优势在最大独立集问题及轻度退极化噪声下仍存在。

摘要:

量子近似优化算法(QAOA)以及一般变分算法的一个挑战是找到好的变分参数,这本身可能是一个NP难问题。最近的工作试图通过为变分参数选择信息充分的猜测来去变分化QAOA。线性斜坡QAOA(LR-QAOA)通过使用受量子绝热算法启发的参数调度来实现这一点。在这项工作中,我们提出了谱隙感知斜坡QAOA(SGIR-QAOA),这是一种新的QAOA变体,它结合了来自绝热哈密顿量的谱隙信息,以QAOA混合哈密顿量作为初始哈密顿量,构建平滑的参数调度。SGIR-QAOA在绝热哈密顿量谱隙较小的地方进行缓慢演化。我们表明,在Grover问题上,SGIR-QAOA在恒定深度下比LR-QAOA有性能提升,并且SGIR-QAOA需要更短的深度来达到相同的最优解概率。然后我们证明这些性能优势扩展到具有潜在实际应用的问题——最大独立集(MIS)问题。最后,我们展示了SGIR-QAOA方法的可扩展性,使用外推的谱隙信息用于无法精确计算谱隙的规模,并表明在轻度退极化噪声下优势似乎仍然存在。

英文摘要:

A challenge with the Quantum Approximate Optimisation Algorithm (QAOA), and variational algorithms in general, is finding good variational parameters, a task which in itself can be NP-hard. Recent work has sought to de-variationalise QAOA by picking well-informed guesses for the variational parameters. The Linear Ramp QAOA (LR-QAOA) achieves this by using parameter schedules inspired by the quantum adiabatic algorithm. In this work, we propose Spectral Gap Informed Ramp QAOA (SGIR-QAOA), a new QAOA variant that incorporates spectral gap information from an adiabatic Hamiltonian, with the QAOA mixer Hamiltonian as the initial Hamiltonian, to construct smooth parameter schedules. SGIR-QAOA performs slow evolution where the spectral gap of the adiabatic Hamiltonian is small. We show that SGIR-QAOA has performance improvements over the LR-QAOA on Grover's problem at constant depth and that SGIR-QAOA requires shorter depths to achieve the same optimal solution probability. We then show that these performance benefits extend to a problem with potential practical applications - the Maximum Independent Set (MIS) problem. Finally, we demonstrate the scalability of the SGIR-QAOA method using extrapolated spectral gap information for scales that the spectral gap cannot be exactly evaluated, and show that the advantage appears to persist under mild depolarising noise.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑