Publications

The nonorthogonal quantum eigensolver (NOQE) is a promising quantum chemistry algorithm that estimates low-lying molecular energies using shallow circuits on near-term devices. We show that overlap thresholding improves the measurement cost of NOQE from O(M^3) to O(M) shots per matrix element, where M is the number of reference states. Source codes for the numerical simulations, based on PySCF and OpenFermion, are available online.

We identify a novel phase transition in the Rabi model subject to bosonic damping and dephasing, where the latter is non-Gaussian. We show that dephasing is a relevant perturbation, leading to a cascade of stability boundaries for bosonic operators.

QUITS is a modular and flexible circuit-level simulator for quantum low-density parity check (QLDPC) codes. QUITS supports several leading QLDPC families, including hypergraph product codes, lifted product codes, and balanced product codes. Available online.

We simulate the spin-boson model, a paradigmatic model of non-Markovian open quantum systems, using the motional modes of trapped ions. The dissipative behavior of the dynamics is captured by applying randomness to the control parameters.

Analog-quantum simulation derived from tracking the evolution of trapped-ion systems holds the potential to simulate molecular quantum dynamics that is beyond the reach of classical-digital strategies. This Review explores the prospects for developing this quantum advantage.

Pulse optimization is not only for quantum gates; it can also be used for calibration (characterization of system parameters). Here we suggest using pulse optimization to characterize the motional-mode parameters (Lamb-Dicke parameters) of trapped ions with higher accuracy, which is necessary for achieving high-fidelity operations on a chain of many ions.

Erasures, or errors with known locations, are more favorable than typical Pauli errors for quantum error correction. We suggest converting physical noise to erasures on trapped ions by using metastable atomic states as qubit states. Then we compare the error-correction performance of metastable and ground qubits under various physical constraints.

To achieve high-fidelity operations on a long ion chain, the motional-mode parameters must be efficiently characterized with high accuracy. We develop and explore physical models that accurately predict both magnitude and sign of the Lamb-Dicke parameters, as well as devise a parallelized and efficient characterization protocol.

We derive and design filter functions of the laser pulses for two-qubit gates on trapped ions, in order to suppress the effects of time-varying fluctuations of the motional-mode frequencies. This improves the experimentally measured gate fidelity from 99.23(7)% to 99.55(7)% in a five-ion chain.

We apply batch optimization, a technique widely used in machine learning, to finding the laser pulses for two-qubit gates on trapped ions. We experimentally verify that the new method improves the gate’s robustness to offsets of the motional-mode frequencies.