---
license: cc-by-nc-4.0
---

### CheXmix: Unified Generative Pretraining for Vision Language Models in Medical Imaging

CheXmix is a unified early-fusion generative model trained on a large corpus of chest X-rays paired with radiology reports