Verbal-R3: Verbal Reranker as the Missing Bridge between Retrieval and Reasoning

ArXi:2605.01399v1 Announce Type: new The conventional Retrieval-Augmented Generation (RAG) paradigm of injecting raw retrieved texts into the Large Language Model (LLM)'s context often results in suboptimal integration of retrieved information. This paper proposes to bridge retrieval results and the LLM's reasoning ability through Verbal Annotations, analytic narratives that explicitly articulate the logical connection between a search query and retrieved contexts.