Ultra-stable and high-performance squeezed vacuum source enabled via artificial intelligence control

Area of Science:

• Quantum optics
• Quantum metrology
• Atomic physics

Background:

• Squeezed states are vital for quantum metrology, surpassing classical precision limits.
• Generating stable, high-performance squeezed light is challenging due to system complexity and quantum fragility.

Purpose of the Study:

• To achieve record-breaking quantum squeezing levels with long-term stability.
• To overcome limitations in optimizing squeezed light generation using traditional models.
• To demonstrate the application of artificial intelligence in enhancing quantum systems.

Main Methods:

• Experimental generation of squeezed light using polarization self-rotation (PSR) in atomic vapor.
• Development of an artificial intelligence control (AIC) system employing deep learning.
• Integration of AIC with the PSR system for adaptive environmental management.

Main Results:

• Achieved a record squeezing level of 4.3 decibels (5.9 decibels lossless).
• Demonstrated long-term stability of the squeezed light generation for several hours.
• Showcased the AIC system's ability to manage complex parameter interferences and adapt to environmental changes.

Conclusions:

• The integrated approach of PSR and AI control significantly advances quantum metrology.
• This work represents a practical step towards the application of quantum sensing and information processing.
• Highlights the synergy between artificial intelligence and fundamental science in overcoming complexity barriers.