1. machine learning magneto-fingerprints in mesoscopic systems

Tomi Ohtsuki

Sophia University, Japan

1 June 2021 Tue 5 pm

                                      IBS Center for Theoretical Physics of Complex Systems (PCS), Administrative Office (B349), Theory Wing, 3rd floor

                                      Expo-ro 55, Yuseong-gu, Daejeon, South Korea, 34126 Tel: +82-42-878-8633                     

At low temperature, the electric conductance of a nano-sized metal exhibits complex but reproducible patterns as a function of magnetic fields.  This is called quantum fingerprints or magneto-fingerprints.  Such complex patterns are due to quantum interference of electrons.  Although the interference pattern carries such microscopic information, it is too complicated for human eyes to understand.  Here we show that machine learning allows us to decipher magneto-fingerprints; a fingerprint pattern in magneto-conductance is shown to be transcribed into spatial images of electron wave functions in a sample by using generative machine learning. The output wave function reveals quantum interference of conduction electrons as well as the positions of scatterers.  Thus by measuring the magneto-conductivity we can obtain the positions of scatterers inside the sample without relying on the miscroscopes.