数统华章2025系列59 Optimized Protocols and Resource Theory of Quantum State Purification

Quantum state purification—the process of extracting higher-fidelity quantum states from multiple copies of noisy quantum states—plays a crucial role in quantum communication and computation. This talk presents a comprehensive investigation of quantum state purification from both optimization and quantum resource theory perspectives. We first introduce a general purification framework to study the fidelity-probability trade-offs in quantum state purification and show a systematic implementation method via block encoding and parameterized quantum circuits. We further study quantum state purification under stabilizer operations that are essential for fault-tolerant quantum computation. While stabilizer operations can enhance fidelity for specific noisy ensembles, we prove they cannot achieve universal purification, thereby establishing the indispensable role of non-stabilizer or magic resources. For distributed quantum systems, we systematically characterize purification under local operations and classical communication (LOCC). We establish no-go theorems demonstrating that no nontrivial two-to-one LOCC purification protocol exists for fundamental state sets, meaning that entangled resources are necessary for state purification. 香港科技大学（广州）信息枢纽人工智能学域副教授，国家高层次青年人才，长期从事量子信息、量子计算、量子机器学习的研究，目前担任国际权威期刊 IEEE Transactions on Information Theory 副主编、Quantum 的编委、中国科学：信息科学和 Chip 的青年编委。于2019年-2023年在百度研究院从事量子计算研究和平台研发，带领团队开发量子机器学习平台“量桨”。他于2018年在悉尼科技大学获得博士学位，获评校长杰出博士论文并入选教育部国家优秀自费留学生奖学金。在 PRL、PRX Quantum、IEEE TIT、IEEE JSAC、CMP、npj Quantum Information、Quantum、NeurIPS、AAAI、ISCA、HPCA、MICRO 等物理、计算机、人工智能领域的权威期刊与会议发表 80 余篇文章。他受邀在量子计算顶会 TQC 作主旨报告（大陆首个）以及在美国物理学年会 APS March Meeting 做邀请报告。他在量子计算方向申请和授权了 50 余项中美专利。他是 QIP、ISIT、AQIS、QCTiP、QTML、AAAI 等国际学术会议的程序委员会委员，并连续多年入选全球前 2%顶尖科学家榜单。