Arxiv paper - Expert-level validation of AI-generated medical text with scalable language models | AI Breakdown | Podwise