Dresden 2026 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
QI: Fachverband Quanteninformation
QI 13: Quantum Control
QI 13.5: Vortrag
Mittwoch, 11. März 2026, 16:00–16:15, BEY/0245
Learning to Steer Quantum Mmany-Body Dynamics with Tree Optimization — Jixing Zhang1, Bo Peng2, Yang Wang1, •Cheuk Kit Cheung1, Guodong Bian3, Andrew Edmonds4, Matthew Markham4, Zhe Zhao2, Durga Dasari1, Ruoming Peng1, Ye Wei2, and Jörg Wrachtrup1 — 13rd Institute of Physics, University of Stuttgart, Allmandring 13, Stuttgart, 70569, Germany — 2Department of Data Science, City University of Hong Kong, Hong Kong, China — 3School of Chemistry, University of Birmingham, B15 2TT, Edgbaston Birmingham, UK — 4Element Six Global Innovation Centre, Fermi Avenue, Harwell Oxford, Didcot, Oxfordshire OX11 0QR, United Kingdom
Achieving practical quantum technologies requires high-quality control over complex quantum systems, but progress is hindered by exponentially growing state spaces and experimental challenges. We present an AI framework that learns to design optimized pulse sequences for many-body spin control, offering a powerful alternative to conventional theory-driven methods. The framework combines tree search, neural network filtering, and numerical simulation guidance to navigate highly nonlinear optimization landscapes using minimal resources. It identifies high-performing, non-intuitive sequences that established methods struggle to find. Experiments in a diamond spin ensemble show the best AI-designed sequences achieved spin coherence times exceeding 200 microseconds, a 100% improvement over state-of-the-art baselines. This work highlights AI's potential to steer complex many-body dynamics, marking a decisive shift toward data-driven sequence design.
Keywords: Dynamical Decoupling; Manybody Physics; Artificial intelligence; Diamond; NV center