Berlin 2024 – wissenschaftliches Programm

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

QI: Fachverband Quanteninformation

QI 23: Quantum Control

QI 23.12: Vortrag

Donnerstag, 21. März 2024, 12:30–12:45, HFT-FT 131

Reinforcement learning entangling operations for spin qubits — •Mohammad Abedi — Forschungszentrum Jülich. Germany

Traditional methods of optimising control pulses rely on the ability to compute gradients of a model of the system dynamics. We investigate reinforcement learning (RL) is a model-free alternative, which optimises entangling operations directly from experience by interacting with a quantum dot spin qubit system. While employing a detailed numerical model of the quantum chip at this point, we explore how the realistically limited observation on quantum systems can be augmented via sequential autoregressive learning with transformer models.

Keywords: spin qubits; quantum dots; reinforcement learning; quantum control; machine learning