Creation of a tweezer array for cold atoms utilizing a generative neural network

TL;DR

This paper introduces a novel method for creating optical tweezer arrays for cold atoms using a generative neural network, enabling faster and more flexible atom trapping with minimal re-optimization.

Contribution

The study presents a new approach combining generative neural networks with spatial light modulators to generate and control optical tweezer arrays efficiently.

Findings

Successfully loaded cold strontium atoms into various tweezer patterns.

Reduced control process time for the SLM, minimizing delays.

Eliminated the need for repeated re-optimization of holograms.

Abstract

Optical tweezers have become essential tools for dynamically manipulating objects, ranging from microspheres or biological molecules to neutral atoms. In this study, we demonstrate the creation of tweezer arrays using a generative neural network, which allows for the trapping of neutral atoms with tunable atom arrays. We have successfully loaded cold strontium atoms into various optical tweezer patterns generated by a spatial light modulator (SLM) integrated with generative models. Our approach shortens the process time to control the SLM wtih minimal time delay, eliminating the need for repeated re-optimization of the hologram for the SLM.