北京师范大学多尺度自旋物理教育部重点实验室
首页» 学术报告

学术报告

Bipartite Reweight-Annealing Framework for Quantum Monte Carlo: Efficient Computation of Entanglement Entropy and Negativity, Off-Diagonal Measurements, and Quantum Magic

报告人|严正 研究员 西湖大学

时 间|2025.12.12 10:00 

 地 点|物理楼 106

报告人简介

Zheng Yan is an Assistant Professor/Researcher at Westlake University. He received the Bachelor Degree from Dalian University of Technology in 2013 and the Ph.D. in Theoretical Physics from Fudan University in 2019. Thereafter he worked as a Postdoctoral Fellow, then a Research Assistant Professor at the University of Hong Kong. In 2023, he joined the Westlake University as a Principal Investigator. His main research areas are theoretical and numerical studies of quantum many-body physics, as well as related quantum simulation/computing and quantum materials cross-disciplinary research.

ABSTRACT

By applying the idea of reweighting in Monte Carlo calculations, we can anneal the partition function to specific reference points, thereby solving for the partition function across the entire annealing parameter path. Furthermore, most quantum observables can be interpreted as ratios of generalized partition functions corresponding to two distinct manifolds of path integral. Leveraging the concept of reweight-annealing, we independently reweight-anneal each generalized partition function to a common reference point. If the ratio of the two partition functions at this reference point can be readily computed, the value of the physical observable across the entire parameter path is obtained. It is important to note that the annealing path is not limited to physically tunable parameters but can also include unconventional variables such as system size or imaginary time. This constitutes a universal framework for quantum Monte Carlo calculations, which we name the bipartite reweight-annealing algorithm. Using this algorithm, we have efficiently computed physical quantities—such as entanglement entropy, entanglement negativity, off-diagonal operator measurements, and quantum magic resource—that were historically challenging to extract via Monte Carlo methods.


E-mail: 202131140010@mail.bnu.edu.cn Web: http://multiscale.bnu.edu.cn
Address: 北京市海淀区新街口外大街19号