Eitan is developing a foundation model for in silico labeling to reduce the amount of paired imaging required, to achieve robustness and generalization across different datasets and organelles.