URL: https://www.desy.de/news/events/index_eng.html/@@events-view
Breadcrumb Navigation
27.06.2023 / 14:30 - 15:30 / Zeuthen, SR5 | Villa
CQTA SeminarFredholm integral equations for the training of shallow neural networks
Patrick Gelß | Zuse Institut Berlin
Ground state and thermal equilibrium properties of local quantum many-body systems can be explored with tensor networks, thanks to their area law entanglement. Highly excited states or out-of-equilibrium setups are much harder. These are therefore the natural problems in which quantum devices can potentially find the earliest advantage. Energy filters allow us to access properties of the system at finite energy densities. They can be efficiently realized by quantum simulators or computers, which simulate the quantum dynamics, combined with classical filtering and sampling. But also replacing the quantum evolution by its classical simulation with tensor networks provides a new tool to classically compute dynamical properties of much larger systems than allowed by other methods. We have recently explored this and other tensor network alternatives to explore quantum many-body systems beyond their equilibrium.
More Information: https://quantum-zeuthen.desy.de/seminar/index_eng.html
