Applications… – Recent work [Phys. Rev. Lett. 131, 110601 (2023)] with my advisor Soonwon Choi and the Endres lab at Caltech studies a useful application of quantum chaos: the estimation of the fidelity between an experimental state and an ideal one. We discover a new universal behavior in quantum dynamics – in a type of observable that is unique to modern quantum devices – and use such universality to benchmark the accuracy which a quantum device performs a desired, complex operation. We initially demonstrated this experimentally in a Rydberg atom array [Nature 613, 468–473 (2023)] (press article) and a superconducting qubit device [Science 379, 6629, 278-283 (2023)], and we subsequently pushed this further to benchmark a large quantum system and even show that it produced highly entangled states [Nature 628, 71-77 (2024)]. This universality is unique to pure quantum states and is spoiled in the presence of noise, which produces a quantum mixed state. This bug can be turned into a feature: by detecting deviations from universality, we can learn some things about the noise in an experiment [Phys. Rev. X 15, 031001 (2025)]. In collaboration with talented statisticians at MIT’s IDSS, we are current extending these techniques to quantify many (hundreds or thousands) of noise parameters from the same datasets.
… and basic science. – These application-driven studies led to a number of discoveries in basic science, to explain why they were working better than we had expected. The quantum-chaotic behavior we had found turned out to be part of a maximum entropy principle governing the way a chaotic system explored its Hilbert space [Phys. Rev. X 14, 041051 (2024)]. Ways this manifests include the phenomena of Hilbert-space ergodicity and deep thermalization, a re-naming of the projected ensembles we had originally discovered both in experiment and in theory [PRX Quantum 4, 010311 (2023)].