EHRAdapt: Adapting Pretrained Language Models to Electronic Health Records with Semantic Priors for Rare Clinical Events
EHRAdapt, an adapter that maps tuples directly into a frozen language model's embedding space, outperforms text-based LLM and count-based baselines, and both pathways improve rare-disease discrimination.
Andre R. Goncalves, Vincent X. Liu, Priyadip Ray
· 0 citations