Transfer function of an asymmetric superconducting Gauss neuron

Area of Science:

• Quantum computing
• Superconducting electronics
• Nonlinear signal processing

Background:

• The Gauss neuron, a nonlinear signal converter, utilizes a transfer function derived from sigmoidal dependence.
• Superconducting Gauss neurons are implementable using two-junction interferometers with inductive shunting.

Purpose of the Study:

• To analyze the impact of three asymmetry types on Gauss neuron transfer functions.
• To explain experimental observations in asymmetric superconducting Gauss neurons.

Main Methods:

• Mathematical analysis of modified equations for asymmetric cases.
• Comparison of transfer function shapes between symmetric and asymmetric configurations.

Main Results:

• Unequal critical currents in Josephson junctions modify the transfer function.
• Asymmetric inductive shunting alters the neuron's signal conversion characteristics.
• Input signal asymmetry also impacts the transfer function shape.

Conclusions:

• The analysis provides a theoretical explanation for experimental features observed in asymmetric Gauss neurons.
• Understanding these asymmetries is crucial for designing reliable superconducting signal converters.