Zhejiang University

报告摘要：

Inspired by the insight of quantum entanglement, tensor network states, such as projected entangled pair states (PEPS), have emerged as a powerful theoretical and numerical tool to characterize correlated systems. They hold the promise to resolve numerous longstanding quantum many-body problems. However, the inherent complexity of PEPS has made the development of efficient, accurate methods a major challenge in the field. In this talk, I will present our recent algorithmic and conceptual advances in PEPS methodology. First, I will introduce our work on integrating PEPS with variational Monte Carlo, and demonstrate its power in resolving the nature of various frustrated spin models. Then, I will discuss our results for the Hubbard model, demonstrating the finite PEPS approach as a practically powerful tool in the post-DMRG age for fermionic systems. Finally, I will introduce a new perspective on tensor networks----tensor network function, which expands the application of tensor networks into new, promising directions.

报告人简介：

刘文渊，2012年四川大学本科毕业，2017年中国科学技术大学博士毕业，之后在香港中文大学、香港大学和加州理工学院从事博士后研究，2024年10月加入浙江大学物理高等研究院，任研究员。长期从事张量网络等量子多体计算方法的发展，研究兴趣包括量子阻挫磁性，强关联电子体系，格点规范场论，量子模拟，机器学习等。

联系人：廖海军（82649377，navyphysics@iphy.ac.cn）

本次活动由中科院青促会物理所小组主办