Invited Talks & Presentations
Here is a selected list of my invited talks, community seminars, and recorded presentations.
Universal Hamming Weight Preserving Variational Quantum Ansatz
Presented at MindSpore Quantum Community.
Abstract: Hamming Weight Preserving (HWP) quantum circuits are vital for particle number conservation and early-stage fault-tolerant error tracking. Via Lie algebra dimensionality analysis, we provide the necessary and sufficient conditions for two-qubit universal HWP operators under fully-connected and adjacent topologies. We verify that only a fraction of universal operators can precisely approximate arbitrary unitaries within sub-spaces. Applied to electronic structures and Fermi-Hubbard models under over-parameterization, our ansatz solves ground-state energies with errors under $10^{-10}\text{ Ha}$ without structural modifications, offering a theoretically rigorous quantum circuit realization for Full Configuration Interaction (FCI).
📺 Watch on Bilibili 📄 Related Paper
Artificial Intelligence and Quantum Computing for Electronic Structures
Presented at the School of Computer Science, Shanghai Jiao Tong University.
Abstract: Driven by the AI4Science paradigm, this talk introduces foundational quantum chemistry concepts alongside our recent breakthroughs at the intersection of machine learning and electronic structures. Moving beyond atom-level predictions, we explore electron wavefunctions via two distinct paradigms: (1) a first-quantized Neural Network Variational Monte Carlo (NN-VMC) algorithm that utilizes Graph Neural Networks (GNNs) for ab initio calculations on unseen chemical structures, and (2) a second-quantized Hamming Weight Preserving (HWP) variational quantum algorithm. The latter achieves ground-state energy computation errors below $$10^{-10}\text{ Ha}$$ for electronic structures and Fermi-Hubbard models, delivering the first theoretically rigorous quantum circuit implementation for Full Configuration Interaction (FCI).
📺 Watch on Bilibili
Quantum Circuit Synthesis and Compilation Optimization: Overview and Prospects
Presented at MindSpore Quantum Community.
Abstract: This talk reviews the critical milestones in quantum circuit logic design and compilation adaptation (e.g., qubit mapping and routing). Recent literature indicates that integrating Artificial Intelligence (AI) significantly scales up optimization precision and efficiency. Building upon this trend, we explore the feasibility of an all-in-one co-design paradigm that merges logical circuit synthesis with hardware-level compilation optimization. By capitalizing on the cognitive capabilities of machine learning, this framework diminishes manual design overhead, enhances execution fidelity, and accelerates the validation of quantum supremacy on physical hardware.
📺 Watch on Bilibili 📄 Related Paper
A Preliminary Exploration of Quantum Machine Learning
Presented at the School of Computer Science, Shanghai Jiao Tong University, 2023.
Abstract: With the steady evolution of NISQ-era quantum hardware, verifying the empirical advantages of Quantum Machine Learning (QML) has emerged as a frontline research frontier. This seminar provides an introductory overview of QML paradigms alongside our early developments tailored for quantum chemistry and combinatorial optimization problems. Furthermore, within the AI4Quantum domain, we discuss our proposed deformable Quantum Architecture Search (QAS) algorithm, which substantially curtails quantum circuit depth to enhance algorithmic execution fidelity on real-world noisy quantum processors.
📺 Watch on Bilibili