Learning temporal embeddings from electronic health records of chronic kidney disease patients

ArXi:2601.18675v2 Announce Type: replace-cross We investigate whether temporal embedding models trained on longitudinal electronic health records can learn clinically meaningful representations without compromising predictive performance, and how architectural choices affect embedding quality. Model-guided medicine requires representations that capture disease dynamics while remaining transparent and task agnostic, whereas most clinical prediction models are optimised for a single task.