Diffusion model for analyzing quantum fingerprints in conductance fluctuation

Abstract

A conditional diffusion model has been developed to analyze intricate conductance fluctuations called universal conductance fluctuations or quantum fingerprints appearing in quantum transport phenomena. The model reconstructs impurity arrangements and quantum interference patterns in nanometals by using magnetoconductance data, providing a novel approach to analyze complex data based on machine learning. In addition, we visualize the attention weights in the model, which efficiently extract information on the non-local correlation of the electron wave functions, and the score functions, which represent the force fields in the wave-function space.