Fetching the paper…

Understand What LLM Needs: Dual Preference Alignment for Retrieval-Augmented Generation · Around