About Alexander Scheinker
I am developing robust adaptive generative AI for time-varying systems by incorporating hard physics constraints and model-independent feedback directly into AI architectures. Examples of some applications are adaptive physics-informed generative diffusion models for generating megapixel resolution virtual diagnostics of the 6D phase space of relativistic charged particle beams in particle accelerators. Another application is generating the 3D shape and phase fields for 3D coherent diffraction imaging (CDI) reconstructions.