Achieving Functional Meta‐Devices by Generalized Meta‐Atom Model for Metasurfaces and Genetic Algorithm (Advanced Optical Materials 10/2024)

Meta-Atom Model for Metasurfaces By pixelating the unit cell of metasurfaces, the machine-learning-based meta-atom model was developed. The model can cover the entire pixel space and simulate any potential meta-atoms operating at any polarization. In article number 2302255, Rui-Xin Wu and co-workers applied the model to design different types of functional meta-devices, including the reflective and the transmissive ones implemented by one or multiple metasurfaces, and a metasurface with one or more meta-atoms. The designs were verified through simulations and experiments. The model significantly improves the design efficiency of the meta-devices.