Dr. Yongxin Yao is a Senior Scientist at Ames National Laboratory and an Adjunct Professor at Iowa State University. His research focuses on quantum computing algorithms, strongly correlated quantum materials, and advanced electronic structure methods. He has led the development of hybrid quantum–classical approaches for correlated electron systems, including adaptive variational quantum algorithms and quantum embedding methods implemented across multiple quantum computing platforms. Dr. Yao also develops scalable computational frameworks integrating AI/ML, high-performance computing, and first-principles many-body techniques for materials discovery and non-equilibrium quantum dynamics. He has authored more than 100 peer-reviewed publications spanning quantum computing, condensed matter theory, ultrafast dynamics, and computational materials science. His work has been supported by the U.S. Department of Energy, and he currently leads projects on quantum computing enhanced simulations of correlated materials.
Presentation Title:
Quantum-Classical Embedding and Adaptive Variational Quantum Algorithms for Correlated Electron Systems
Presentation Abstract:
Simulating correlated electron systems on current and near-term quantum hardware remains challenging due to limited qubit counts, circuit depth, and noise. Quantum embedding approaches provide a promising route by mapping complex lattice problems onto effective impurity models that can be treated with reduced quantum resources. In this work, we develop a quantum-classical framework based on the ghost Gutzwiller approximation for quantum-enhanced simulations of correlated materials, including both equilibrium and dynamical properties. We combine this framework with adaptive variational quantum algorithms to study representative strongly correlated models and analyze the associated quantum resource requirements. For the infinite-dimensional Hubbard model, increasing the number of ghost modes from 3 to 5 leads to circuit depths ranging from 16 to 104. Simulations with realistic noise models reveal substantial degradation in the spectral properties, motivating the implementation of the Iceberg quantum error-detection code, which achieves up to 40% error reduction. Finally, density matrices, spectral functions, and dynamical observables are benchmarked on IBM and Quantinuum quantum hardware using multiple levels of error-mitigation techniques.