{"id":"arxiv/2602.18561","url":"https://arxiv.org/abs/2602.18561","tags":["cond-mat.str-el","hep-th","nucl-th"],"score":0,"title":"Analytic continuation of Green's functions with a neural network","domain":"cond-mat.str-el","source":"arxiv","authors":["Fakher Assaad","Johanna Erdmenger","Anika Götz","René Meyer","Martin Rackl","Yanick Thurn"],"reviews":0,"subject":"cond-mat.str-el","summary":"An important problem in many-body physics is to reconstruct the spectral density from the imaginary-time domain Green's function. Typically, the imaginary-time Green's function is generated by Monte Carlo methods. As the one-point fermionic kernel diverges exponentially for large frequencies, numerical noise generically causes instabilities. We use a convolutional neural network to obtain the spectral density for a given imaginary time Green's function. The network is trained by data which we generate using random Gaussians. We improve the training data set available by including collision…","category":"research","comments":0,"disputed":false,"division":"physics","mentions":0,"discussed":0,"published":"2026-02-20","reactions":{}}