IMRNNs: An Efficient Method for Interpretable Dense Retrieval via Embedding Modulation

2026-01-27

AI Summary

This work addresses the limited interpretability of black-box dense retrieval models, which obscures the semantic interaction mechanisms between queries and documents. To this end, we propose the IMRNNs framework, which introduces a lightweight bidirectional dynamic embedding modulation mechanism—the first of its kind—that dynamically conditions document embeddings on the query during inference. Furthermore, IMRNNs iteratively refines query representations through feedback from initial retrieval results, enabling bidirectional semantic alignment between queries and documents. This approach enhances both retrieval effectiveness and model interpretability without incurring significant computational overhead. Evaluated across seven benchmark datasets, IMRNNs achieves average improvements of 6.35% in nDCG, 7.14% in recall, and 7.04% in MRR, substantially outperforming state-of-the-art dense retrieval methods.

Technology Category

Machine Learning: Large Multimodal Models (LMMs)Data Mining & Knowledge Management: Conversational Systems for Recommendation & RetrievalNatural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models

Application Category

Search and Retrieval-Augmented AI: Large language models for searchUser Modeling, Personalization and Recommendation: Fairness-aware retrieval and rankingSemantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Interpretability in black-box dense retrievers remains a central challenge in Retrieval-Augmented Generation (RAG). Understanding how queries and documents semantically interact is critical for diagnosing retrieval behavior and improving model design. However, existing dense retrievers rely on static embeddings for both queries and documents, which obscures this bidirectional relationship. Post-hoc approaches such as re-rankers are computationally expensive, add inference latency, and still fail to reveal the underlying semantic alignment. To address these limitations, we propose Interpretable Modular Retrieval Neural Networks (IMRNNs), a lightweight framework that augments any dense retriever with dynamic, bidirectional modulation at inference time. IMRNNs employ two independent adapters: one conditions document embeddings on the current query, while the other refines the query embedding using corpus-level feedback from initially retrieved documents. This iterative modulation process enables the model to adapt representations dynamically and expose interpretable semantic dependencies between queries and documents. Empirically, IMRNNs not only enhance interpretability but also improve retrieval effectiveness. Across seven benchmark datasets, applying our method to standard dense retrievers yields average gains of +6.35% nDCG, +7.14% recall, and +7.04% MRR over state-of-the-art baselines. These results demonstrate that incorporating interpretability-driven modulation can both explain and enhance retrieval in RAG systems.

Problem

• Interpretability
• Dense Retrieval
• Semantic Interaction
• Retrieval-Augmented Generation
• Black-box Models

Innovation

• interpretable retrieval
• embedding modulation
• dense retrieval
• bidirectional adaptation
• RAG