Decide Then Retrieve: A Training-Free Framework with Uncertainty-Guided Triggering and Dual-Path Retrieval

2026-01-07

arXiv.org

AI Summary

This work addresses the limitations of existing retrieval-augmented generation (RAG) approaches, which lack selectivity in retrieval triggering and rely on single-path evidence, often introducing noise and constraining performance. The authors propose a training-free framework that leverages the generation uncertainty of large language models to adaptively determine whether to invoke retrieval. Furthermore, they introduce a dual-path mechanism that separately handles sparse and ambiguous queries, enabling more precise information selection and fusion. Evaluated on five open-domain question answering benchmarks, the method significantly outperforms standard RAG and strong baselines, achieving consistent gains in Exact Match (EM) and F1 scores while substantially reducing redundant retrieval calls.

Technology Category

• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Natural Language Processing: Question Answering
• Reasoning under Uncertainty: Other Foundations of Reasoning under Uncertainty

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs

Abstract

Retrieval-augmented generation (RAG) enhances large language models (LLMs) by incorporating external knowledge, but existing approaches indiscriminately trigger retrieval and rely on single-path evidence construction, often introducing noise and limiting performance gains. In this work, we propose Decide Then Retrieve (DTR), a training-free framework that adaptively determines when retrieval is necessary and how external information should be selected. DTR leverages generation uncertainty to guide retrieval triggering and introduces a dual-path retrieval mechanism with adaptive information selection to better handle sparse and ambiguous queries. Extensive experiments across five open-domain QA benchmarks, multiple model scales, and different retrievers demonstrate that DTR consistently improves EM and F1 over standard RAG and strong retrieval-enhanced baselines, while reducing unnecessary retrievals. The code and data used in this paper are available at https://github.com/ChenWangHKU/DTR.

Problem

• Retrieval-augmented generation
• uncertainty-guided triggering
• dual-path retrieval
• unnecessary retrieval
• external knowledge integration

Innovation

• Retrieval-Augmented Generation
• Uncertainty-Guided Retrieval
• Dual-Path Retrieval
• Training-Free Framework
• Adaptive Information Selection