ModernBERT + ColBERT: Enhancing biomedical RAG through an advanced re-ranking retriever

ICLR 2026 Conference Submission13863 Authors

Keywords: Medical Imaging, LLM, RAG, Agents

TL;DR: This paper introduces a two-stage retrieval system with ModernBERT and ColBERTv2 that boosts biomedical RAG accuracy and efficiency, achieving state-of-the-art MIRAGE results with low computational cost.

Abstract:

Retrieval-Augmented Generation (RAG) is a powerful technique for enriching Large Language Models (LLMs) with external knowledge, allowing for factually grounded responses, a critical requirement in high-stakes domains such as healthcare. However, the efficacy of RAG systems is fundamentally restricted by the performance of their retrieval module, since irrelevant or semantically misaligned documents directly compromise the accuracy of the final generated response. General-purpose dense retrievers can struggle with the nuanced language of specialised domains, while the high accuracy of in-domain models is often achieved at prohibitive computational costs. In this work, we aim to address this trade-off by developing and evaluating a two-stage retrieval architecture that combines a lightweight ModernBERT bidirectional encoder for efficient initial candidate retrieval with a ColBERTv2 late-interaction model for fine-grained re-ranking.
We conduct comprehensive evaluations of our retriever module performance and RAG system performance in the biomedical context, fine-tuning the IR module using 10k question-passage pairs from PubMedQA. Our analysis of the retriever module confirmed the positive impact of the ColBERT re-ranker, which improved Recall@3 by up to 4.2 percentage points compared to its retrieve-only counterpart. When integrated into the biomedical RAG, our IR module leads to a state-of-the-art average accuracy of 0.4448 on the five tasks of the MIRAGE question-answering benchmark, outperforming strong baselines such as MedCPT (0.4436). Our ablation studies reveal that this performance is critically dependent on a joint fine-tuning process that aligns the retriever and re-ranker; otherwise, the re-ranker might degrade the performance. Furthermore, our parameter-efficient system achieves this result with an indexing speed over 7.5 times faster than leading baselines, providing a practical pathway for developing trustworthy biomedical RAG systems. Our implementation is available at: \url{https://anonymous.4open.science/r/biorag-MC-9F3D/}

Primary Area: applications to physical sciences (physics, chemistry, biology, etc.)

Submission Number: 13863