标题： 量子李雅普诺夫张量重构 显示英文标题

标题： Quantum Liouvillian Tomography

摘要： 对近期量子计算平台的表征需要能够捕捉和量化耗散效应的能力。 这是一个本质上具有挑战性的任务，因为这些效应是多方面的，涵盖了从马尔可夫到强非马尔可夫动力学的广泛范围。 我们引入了量子李群张量重构（QLT），一种用于捕捉和量化连续时间量子动力学中非马尔可夫效应的协议。 该协议利用基于梯度的量子过程重构来重建动力学映射，并利用泡利字符串概率分布的导数进行回归，以提取支配动力学的李群算子。 我们使用合成数据对该协议进行基准测试，并量化其在恢复两量子比特系统中的哈密顿量、跳跃算符和耗散率方面的准确性。 最后，我们将QLT应用于分析超导量子平台上实现的空转两量子比特系统的演化，以提取哈密顿量和耗散成分的特性，并因此检测出固有的非马尔可夫动力学。 我们的工作引入了第一个能够从实验数据中检索通用开放量子演化的生成器的协议，从而实现了对近期量子计算平台中多体非马尔可夫效应更精确的表征。 显示英文摘要

摘要： Characterization of near-term quantum computing platforms requires the ability to capture and quantify dissipative effects. This is an inherently challenging task, as these effects are multifaceted, spanning a broad spectrum from Markovian to strongly non-Markovian dynamics. We introduce Quantum Liouvillian Tomography (QLT), a protocol to capture and quantify non-Markovian effects in time-continuous quantum dynamics. The protocol leverages gradient-based quantum process tomography to reconstruct dynamical maps and utilizes regression over the derivatives of Pauli string probability distributions to extract the Liouvillian governing the dynamics. We benchmark the protocol using synthetic data and quantify its accuracy in recovering Hamiltonians, jump operators, and dissipation rates for two-qubit systems. Finally, we apply QLT to analyze the evolution of an idling two-qubit system implemented on a superconducting quantum platform to extract characteristics of Hamiltonian and dissipative components and, as a result, detect inherently non-Markovian dynamics. Our work introduces the first protocol capable of retrieving generators of generic open quantum evolution from experimental data, thus enabling more precise characterization of many-body non-Markovian effects in near-term quantum computing platforms.