Constructive Quantum Interference at the Edge of Quantum Chaos and Application to Hamiltonian Learning

I will present the observation of constructive quantum interference at the edge of quantum chaos on a 103-qubit digital superconducting processor, performing at a scale that exceeds the simulation capacity of the best-known classical algorithms. This interference is a ubiquitous property of a family of out-of-time-order correlators (OTOCs and OTOC2s) that require multiple inversions of the arrow of time. These quantities remain remarkably sensitive to the microscopic features of the underlying dynamics at long times. I will discuss how this property can be utilized for Hamiltonian learning and describe experimental observations for resolving molecular structure in NMR experiments.