非平衡量子多体系统中的受限典型性
Restricted typicality in non-equilibrium quantum many-body systems
浏览论文内容
中文总结 AI 辅助
本文提出全局受限典型性:在守恒荷输运主导的量子多体系统中,仅用慢模式约束最大熵系综即可重构全局态的纠缠熵等非线性性质,推广了典型性范式。
中文摘要 AI 辅助
刻画一般量子多体系统的时间演化是一个根本性挑战,因为表示精确状态需要指数级增长的计算资源。虽然流体力学和统计力学通过预测局域可观测量的期望值成功地简化了这一任务,但这些宏观框架无法提供关于量子态非线性特征的任何信息。在本工作中,我们证明,对于表现出时间尺度分离、且动力学由守恒荷输运主导的系统,这些丢失的信息可以被系统地恢复。通过仅用系统的慢模式约束最大熵Scrooge系综,我们精确地重构了全局时间演化态的复杂非线性量子性质,包括半链纠缠熵和计算基下的参与熵。我们的结果推广了正则典型性范式,揭示了一个受流体动力学瓶颈限制的系统在希尔伯特空间的一个动态受限子流形内表现得像一个典型纯态——我们将这一现象称为全局受限典型性。
英文摘要
Characterizing the time evolution of generic quantum many-body systems is a fundamental challenge, as representing the exact state requires exponentially scaling computational resources. While hydrodynamics and statistical mechanics successfully simplify this task by predicting the expectation values of local observables, these macroscopic frameworks provide no information about nonlinear characteristics of the quantum state. In this work, we demonstrate that for systems exhibiting a timescale separation, with dynamics governed by the transport of conserved charges, this lost information can be systematically recovered. By constraining the maximum-entropy Scrooge ensemble solely by the system's slow modes, we accurately reconstruct complex, nonlinear quantum properties of the global time-evolved state, including the half-chain entanglement entropy and the participation entropy in the computational basis. Our results generalize the paradigm of canonical typicality, revealing that a hydrodynamically bottlenecked system behaves as a typical pure state within a dynamically restricted submanifold of the Hilbert space - a phenomenon we term global restricted typicality.
发表机构
- Szkoła Doktorska Nauk Ścisłych i Przyrodniczych, Uniwersytet Jagielloński(雅盖隆大学自然科学博士学院)
- Instytut Fizyki Teoretycznej, Wydział Fizyki, Astronomii i Informatyki Stosowanej, Uniwersytet Jagielloński(雅盖隆大学物理、天文学与应用信息学系理论物理研究所)
- Barcelona Supercomputing Center(巴塞罗那超级计算中心)
- Mark Kac Complex Systems Research Center, Jagiellonian University in Krakow(克拉科夫雅盖隆大学马克·卡克复杂系统研究中心)
机构由 AI 辅助整理,请以论文原文为准。