Recent advancements in continuous-variable (CV) quantum systems are opening up growing opportunities for applications in both quantum learning and quantum simulation. In this talk, I will first discuss machine-learning approaches to multi-mode quantum tomography, enabling efficient reconstruction and characterization of high-dimensional CV states beyond conventional techniques. I will then discuss recent results of provably efficient Hamiltonian learning protocols for continuous-variable and hybrid quantum computers, achieving Heisenberg-limited precision. Next, I will turn to quantum simulation with continuous variables, highlighting how bosonic modes provide a natural language for representing gauge fields in quantum field theories. In particular, I will present progress in simulating (2+1)-dimensional quantum electrodynamics, where rich topological phases emerge. Overall, these developments point toward the potential of CV quantum processors as a useful framework for advancing learning and simulation capabilities in complex quantum systems.

罗迪，清华大学物理系副教授。此前，他曾任美国加州大学洛杉矶分校（UCLA）助理教授。2021年至2024年间，他担任美国人工智能与基本相互作用研究所IAIFI Fellow，同时是麻省理工学院（MIT）理论物理中心与哈佛大学（Harvard）物理系博士后。2021年，他在伊利诺伊大学厄巴纳-香槟分校（UIUC）获得数学与物理双硕士学位及物理学博士学位。2016年，他毕业于香港大学，获物理学与数学双学士学位。罗迪博士曾在Google Quantum AI, Vector Institute 和 Flatiron Institute 研究实习。他的研究聚焦在AI+Science和量子计算，包括发展AI和量子算法用于量子材料、高能物理和量子信息的科学模拟和发现，以及利用量子物理和统计物理发展构建AI理论和模型。