Fig. 1: Concept of quantum geometric decoder.
From: Deciphering quantum fingerprints in electric conductance
a, b A schematic illustration of a magneto-conductance measurement in a small metal sample (a) and its magnified view (b). The green stripe pattern describes electron wave function intensity (WI) in the sample with defects. c Conductance change \({\Delta}G\) for a nanowire sample. e, h, and a are the elementary charge, the Plank constant, and the lattice constant in the calculation model, respectively. \({\phi }_{0}=h/e\) is the magnetic flux quantum. d Concept of Quantum Geometric Decoding based on a deep neural network. First, the networks (A) and (B) compress the WI images into the latent space. Then, the networks (C) and (B) output a geometry image including the sample shape, defect distribution, and WI information from the input of the magneto-conductance.
