报告人：李俊 教授（深圳大学）

时 间：2026年11月6日（周五）14:30

地 点：良乡校区 信创A 558

摘要：Unitary designs are vital for investigating quantum randomness in both theory and practice. Existing studies focused on discrete unitary designs such as Pauli or Clifford group. We initiate the study of continuous unitary designs (CUD), i.e., continuous unitary path that can emulate Haar-averaging statistics of the unitary group. This report consists of three aspects of research.

(1) Theory. We develop systematical methods for constructing CUDs. Our theory offers novel insights into quantum randomness from geometric, topological, and combinatorial perspectives, which have not been explored before.

(2) Application. We demonstrate CUD applications for engineering noisy quantum dynamics, including universally robust control, quantum memory, noise tailoring and gate benchmarking.

(3) Experiment. We propose CUD-based dynamical decoupling (DD) for removing unwanted interactions of a qubit network. These protocols show advantages over conventional DD protocols. As a remarkable result, in bounded-strength 2-local DD, our CUD sequence length scales as O(N), compared with the previous scaling of O(N log N). We performed initial experimental tests on a superconducting-qubit cloud platform, confirming improved performance in state protection on one to five qubits.

简历：李俊，深圳大学教授、博士生导师。2015年毕业于中国科学技术大学近代物理系，获物理学博士学位；2018年加入南方科技大学，2023年进入深圳大学。长期从事量子系统控制研究，主要方向包括：（1）量子控制的理论方法；（2）量子控制在量子计算与模拟、量子精密测量中的应用。主要成果包括：提出量子学习的参数移位法则，发展多目标鲁棒量子控制方法，提出连续幺正设计理论，以及完成量子混沌、伪随机生成器、含噪量子传感等多个量子信息方案的实验实现。在国际物理期刊发表论文40多篇，主持了国家科技重大专项课题、广东省量子专项重大项目等科研项目共8项。

联系方式：bozhang_quantum@bit.edu.cn

邀请人： 张博

网 址：http://physics.bit.edu.cn/

承办单位：物理学院