SKM 2023 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
TT: Fachverband Tiefe Temperaturen
TT 2: Focus Session: Physics Meets ML I – Machine Learning for Complex Quantum Systems (joint session TT/DY)
TT 2.7: Vortrag
Montag, 27. März 2023, 12:15–12:30, HSZ 03
Simulating spectral functions of two-dimensional systems with neural quantum states — •Tiago Mendes Santos1, Markus Schmitt2, and Markus Heyl1 — 1University of Augsburg, Augsburg, Germany — 2Forschungszentrum Jülich, Jülich, Germany
Spectral functions are key tools to characterize and probe condensed matter systems. Simulating such quantities in interacting two-dimensional quantum matter is, however, still an outstanding challenge. This work presents a numerical approach to simulate spectral functions using Neural Quantum States. As the key aspect, our scheme leverage the flexibility of artificial-neural-network wave functions to access spectral properties by simulating the dynamics of localized excitations with the time-dependent variational Monte Carlo. For demonstration, we study the dynamical structure factor (DSF) of models describing two-dimensional quantum phase transitions, namely, the quantum Ising and a square-lattice Rydberg Atom Arrays model in a regime of parameters relevant to quantum simulators. When combined with deep network architectures whose number of variational parameters increase at a mild polynomial expense with the number of spins, we showcase that our approach reliably describes the DSF for unprecedented system sizes and time scales.