BEGIN:VCALENDAR
VERSION:2.0
PRODID:Icfo
X-PUBLISHED-TTL:P1W
BEGIN:VEVENT
UID:69d4ac8e5d499
DTSTART:20220920T130000Z
SEQUENCE:0
TRANSP:OPAQUE
LOCATION:University of Warsaw
SUMMARY:ICFO | ANNA DAWID
CLASS:PUBLIC
DESCRIPTION:Quantum many-body physics poses a substantial computational cha
 llenge resulting from the exponential growth of the wave function complexi
 ty and many non-trivial correlations encoded in it. Studying many-body sys
 tems is thus a demanding quest that is being tackled via various methods. 
 The research described within this thesis concerns\ntwo parallel approache
 s that are gaining the attention of the scientific community: quantum simu
 lations with ultracold molecules and interpretable machine learning.\nThe 
 first research path is a detailed analysis of the ultracold system of two 
 interacting molecules in a one-dimensional trap. By comparing with the two
 -atom system in a harmonic trap\, we identify differences in spectra and r
 eactions to the external fields introduced by the molecular character of t
 he system\, i.e.\, rotational levels\, anisotropic shortrange interactions
 \, and stronger dipolar interactions. Exactly these richer properties of m
 olecules could allow for discovering new exotic phases of matter and simul
 ating phenomena that are inaccessible for the physics of ultracold atoms. 
 Inspired by materials with both electric and magnetic orders\, in the next
  step\, we focus on the interplay of the electric and magnetic properties 
 of the two-body molecular system\, analyze magnetization diagrams\, and st
 udy the quench dynamics.\nAlternatively\, quantum many-body problems can b
 e solved via numerical methods. Among them\, machine learning algorithms a
 re gaining significant momentum. However\, so far\, they have mostly enabl
 ed only the recovery of known results (but at much lower computational cos
 t). Moreover\, we usually lack the understanding of how the machine solves
  the problem at hand. Therefore\, we propose a way to combine the efficien
 cy of neural networks with Hessian-based interpretability and reliability 
 methods like influence functions. In principle\, these universal and model
 -independent tools allow to unravel the logic hidden in the machine and th
 us increase the chance to understand the physics of the problem. We show t
 heir power on the fundamental one-dimensional Fermi-Hubbard model and on t
 he experimental data obtained from the Floquet realization of the topologi
 cal two-dimensional Haldane model.\n&nbsp\;\nThesis Director:&nbsp\;Prof D
 r. Maciej Lewenstein and Prof Dr. Michał Tomza
DTSTAMP:20260407T070446Z
END:VEVENT
END:VCALENDAR