操作依赖性量子记忆效应及非马尔科夫过程恢复的实验演示数据

主要面向量子物理基本问题的研究，受到国家重点研发计划项目2017YFA0304102支持，记忆效应在量子信息科学中普遍存在，理解和控制记忆效应对于发展量子技术至关重要。系统和环境的耦合导致开放量子系统演化会展现记忆效应。然而，由于测量坍缩效应，此前人们对量子记忆效应的研究局限在单步演化，即初态制备后进行演化最后进行测量。对于多步演化过程，此前度量量子记忆效应的判据比如BLP判据和RHP判据等，无法消除中间测量对系统的测量塌缩效应，因此不能用来研究多步演化的记忆效应。我们利用过程矩阵模型构造多比特模拟器解决这一问题，通过将多体系统空间关联模拟多步演化的时间关联，我们研究了多步演化的量子记忆效应，证明了量子记忆长度的操作依赖性和量子记忆强度的可擦除性。这个工作开启了多时间节点量子非马尔科夫性的实验研究，充分展示了超出之前两时间节点设置的高阶量子记忆效应特征。数据量为96KB。

This work focuses on fundamental research in quantum physics, supported by the National Key R&D Program of China under Grant No. 2017YFA0304102. Memory effects are ubiquitous in quantum information science, and understanding and controlling them is crucial for the development of quantum technologies. The coupling between a quantum system and its environment gives rise to memory effects in the evolution of open quantum systems. However, due to the measurement collapse effect, previous studies of quantum memory effects were limited to single-step evolution: after preparing the initial state, the system undergoes evolution followed by a final measurement. For multi-step evolution processes, previous criteria for quantifying quantum memory effects, such as the BLP criterion and RHP criterion, cannot eliminate the measurement collapse induced by intermediate measurements on the system, thus failing to study memory effects in multi-step evolution. To address this issue, we constructed a multi-qubit simulator using the process matrix model. By simulating the temporal correlations of multi-step evolution via the spatial correlations of a many-body system, we investigated quantum memory effects in multi-step evolution, and demonstrated the operation-dependent nature of quantum memory length and the erasability of quantum memory strength. This work initiates experimental studies of quantum non-Markovianity at multiple time nodes, fully showcasing the characteristics of high-order quantum memory effects beyond the previous two-time-node setup. The dataset has a size of 96 KB.