RESEARCH

Supervised Fine-tuning with Synthetic Rationale Data Hurts Real-World Disease Prediction

ArXiv cs.AI · Wed, 10 Jun 2026 04:00:00 GMT

arXiv:2606.10279v1 Announce Type: new Abstract: Supervised fine-tuning with synthetic rationale data is widely assumed to improve language model performance on clinical prediction tasks by teaching models not just what to predict but why. We test this assumption on five-year Alzh

Read original source Discuss with A.S.I.S