Beware of entropy phase transition! How to make quantum denoising successful?
POSTER
Abstract
Quantum autoencoders can help to generate denoised entanglement on a noisy neural network. However noise outweighs the process if it gets too strong. [1]
We show that the flow of Rényi entropy in a subnet can perform as a measure whose phase transition determines denoising success or failure. Moreover, we rely on the study to implement a deformation of the network that results in improved tolerance against heavier noises.
References:
[1] D. Bondarenko and P. Feldmann, “Quantum autoencoders to denoise quantum data,” Phys. Rev. Lett., vol. 124, no. 13, p. 130502, 2020.
We show that the flow of Rényi entropy in a subnet can perform as a measure whose phase transition determines denoising success or failure. Moreover, we rely on the study to implement a deformation of the network that results in improved tolerance against heavier noises.
References:
[1] D. Bondarenko and P. Feldmann, “Quantum autoencoders to denoise quantum data,” Phys. Rev. Lett., vol. 124, no. 13, p. 130502, 2020.
Publication: J. Pazem, M. H. Ansari, "Quantum state denoising on elongated autoencoders", To be published.
Presenters
-
Joséphine Pazem
IQI, RWTH Aachen University and PGI, Forschungszentrum Jülich, Germany
Authors
-
Joséphine Pazem
IQI, RWTH Aachen University and PGI, Forschungszentrum Jülich, Germany
-
Mohammad H Ansari
PGI, Forschungszentrum Jülich and IQI, RWTH Aachen University, Germany, Forschungszentrum Jülich GmbH