Post-training & Searching RNA Foundation Models Diagnosing diversity collapse and using structural feedback to turn pretrained RNA generation into exact design. Why Does Generative Learning Help Deterministic Prediction? Understanding how target-side molecular signals reshape pathology representations. Structure-Preserving Generative Modeling for Spatial Omics BioFlow keeps histology-conditioned gene-expression trajectories within biologically valid support. Learning Shared and Modality-Specific Structure in Spatial Multi-Omics SpaMV disentangles shared biological structure from signals that are private to each spatial omics modality.