The continual learning (CL) paradigm aims to enable neural networks learn tasks continually in a sequential fashion. fundamental challenge this is catastrophic forgetting previously learned when the model optimized for new task, especially their data not accessible. Current architectural-based methods aim at alleviating problem but expense of expanding capacity model. Regularization-based maint...

We construct optical beams in free space with robust skyrmionic structures their polarization fields, both the electric spin vector for near-circular fields and direction near-linear Bloch (spiral) N\'eel (hedgehog) textures. These are made possible by spin-orbit coupling of tightly-focused nonparaxial optics as applied to higher-order Full-Poincar\'e beams, well standing-wave configurations co...