Dresden 2020 – wissenschaftliches Programm

Die DPG-Frühjahrstagung in Dresden musste abgesagt werden! Lesen Sie mehr ...

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

DY: Fachverband Dynamik und Statistische Physik

DY 66: Many-body Quantum Dynamics II

DY 66.5: Vortrag

Freitag, 20. März 2020, 11:00–11:15, HÜL 186

Reinforcement Learning for Digital Quantum Simulation — •Adrien Bolens and Markus Heyl — Max Planck Institute for the Physics of Complex Systems, Dresden, Germany

Digital quantum simulations (DQS) are one of the most appealing applications of a quantum computer. In theory, the unitary time-evolution of any spin-type Hamiltonian can be encoded in a quantum computer with arbitrary precision. In practice, however, unitary gates entangling different qubits are an important source of error. We use tailored reinforcement learning techniques to optimally generate DQS of collective quantum spin systems, such as the long-range Ising model and the XX model. We show that for a fixed number of entangling quantum gates, the DQS error found is systematically lower than the Trotter error, and decay much more slowly with the system size. In addition, our method let us generate efficient DQS even for models without a well-defined Trotterization in the quantum simulator. We discuss the implications of our algorithm for the implementation of DQS in trapped-ion quantum simulators.