Quantum computing explores the fundamental connections between computation, information processing, and quantum mechanics. On one hand, one asks how we can harness the unique behaviors of quantum mechanics, such as superposition and entanglement, for computation. On the other hand, one explores how computation can help us better understand our physical world. These are the central questions in this field, having significant implications for technological advancements, materials science, and fundamental research in computer science, mathematics, and natural sciences. The main goal of my research is to advance our understanding of these questions by developing new computational and theoretical tools.
Over the last few years, quantum hardware has seen huge advancements, not just in the number of qubits but also in their quality. Noisy intermediate-scale quantum (NISQ) devices, consisting of tens to hundreds of high-quality physical qubits, are becoming increasingly accessible. We’ve transitioned from the era of small, experimental quantum devices to the era of quantum utility, where quantum computers are becoming useful computational tools for scientific research. These advances have opened up new possibilities and challenges, making this a particularly exciting time for doing quantum computing research.
One of my current research interests is developing quantum-utility algorithms that are inherently noise-robust and scalable, designed not only to maximize the potential of current quantum systems but also to pave the way for error-corrected quantum computers of the future.
Measurement-Induced Entanglement Phase Transitions in Variational Quantum Circuits (with R. Wiersema, J. F. Carrasquilla and Y. B. Kim), SciPost Phys., 14, 147, 2023.
Exploring Entanglement and Optimization within the Hamiltonian Variational Ansatz (with R. Wiersema, Y. de Sereville, J. F. Carrasquilla, Y. B. Kim and H. Yuen, PRX Quantum 1, 020319, 2020.
Despite the tremendous advancements in quantum hardware, demonstrating practical quantum advantages on these devices remains a significant challenge. At the same time, fully fault-tolerant quantum (FFTQ) computing, which relies on quantum error correction, is still far from realization due to the immense demand for high-quality physical qubits. Evolving in the regime between NISQ and FFTQ computing, early fault-tolerant quantum (EFTQ) computers––featuring small-scale, limited quantum error correction––offer great promise for demonstrating practical quantum advantages much sooner, potentially within the next decade. New algorithmic frameworks that can take full advantage of these emerging hardware capabilities are essential to realizing this potential.
One of my current research focuses is the development of computational frameworks and tools to bridge the gap between NISQ and EFTQ computation, with the ultimate goal of achieving practical quantum advantages.
Ancilla-Free Quantum Protocol for Thermal Green’s Functions (with C. Yi), 2025
Quantum Phase Estimation by Compressed Sensing (with C. Yi and J. Takahashi), Quantum 8, 1579, 2024.
In quantum systems, exotic quantum effects typically emerge at or near absolute zero temperature. But how complex are these low-temperature quantum states? This question stands as one of the most challenging and significant in condensed matter physics and theoretical computer science, with profound theoretical implications, such as the Quantum Probabilistically Checkable Proofs (PCP) Conjecture, and practical applications, including the pursuit of room-temperature superconductivity. Interestingly, computational approaches, such as tensor networks, optimization, machine learning, and computational complexity theory, provide powerful tools for exploring these questions.
I’m interested in developing 1) novel computational tools for probing quantum many-body models, such as SDP-based and machine learning (ML)-driven approaches, and 2) theoretical tools for understanding the computational complexity of Local Hamiltonian problems (the quantum analogue of MAX-SAT), such as the Quantum MaxCut problem.
The interplay between AI and quantum computing has emerged as one of the most dynamic frontiers in modern computational science. On one hand, quantum algorithms have been proposed to accelerate key ML primitives such as linear algebra routines and kernel estimation, potentially leading to quantum-accelerated learning frameworks. On the other hand—and more relevant to my research program—the reverse direction AI-enhanced quantum computing has shown remarkable potential in addressing the unique challenges of NISQ and emerging fault-tolerant quantum devices. These approaches treat the quantum processor as part of a closed learning loop, in which data, gradients, and feedback flow dynamically between quantum and classical components to optimize circuit parameters, gate configurations, and control strategies.
Although rapid progress has been made in exploring AI-driven approaches, many fundamental and practically relevant challenges remain open, such as developing general frameworks for finding optimal or near-optimal quantum circuits for measurement protocols (a core task in quantum computing), as well as scalable learning of quantum states and processes, data-efficient and accurate noise characterization, large-scale error mitigation, and efficient decoding algorithms for quantum error correction. One of my current research focuses is to make progress toward addressing these challenges by developing principled, learning-driven approaches based on modern ML paradigms.