A superconducting adiabatic neuron in a quantum regime

Area of Science:

• Quantum computing
• Artificial intelligence
• Computational neuroscience

Background:

• Hybrid classical-quantum systems offer novel computational paradigms.
• Artificial neural networks are foundational to machine learning.
• Quantum dynamics can influence classical system behavior.

Purpose of the Study:

• To investigate the quantum dynamics of an adiabatic neural cell within a perceptron artificial neural network.
• To analyze a hybrid system where a quantum co-processor adjusts a classical neural network's configuration.
• To determine the conditions for achieving a sigmoid activation function in such a quantum-classical hybrid neuron.

Main Methods:

• Analytical and numerical studies were employed.
• The research considered non-adiabatic processes and dissipation effects.
• Quantum coherent oscillations were analyzed for their smoothing due to these factors.

Main Results:

• The study identified specific conditions under which the neural cell exhibits the desired sigmoid activation function.
• Non-adiabatic processes and dissipation were shown to smooth quantum coherent oscillations.
• The dynamics of adiabatic neural cells in a quantum regime were characterized.

Conclusions:

• The findings provide insights into the behavior of hybrid quantum-classical neural networks.
• Achieving a functional sigmoid activation in a quantum-assisted neuron is feasible under certain conditions.
• This research contributes to the development of quantum-enhanced artificial intelligence.