用数学方法选更优检索结果，让大模型回答更准更不重复。
SMART-RAG: Selection using Determinantal Matrices for Augmented Retrieval
- 基于确定性矩阵的无监督筛选机制，兼顾相关性、多样性与冲突避免。
- 在多个数据集上显著提升问答准确率，优于现有无监督方法。
- 无需训练，适配任意大模型，适合追求高效精准检索的开发者。
检索增强生成（RAG）通过引入外部信息显著提升了大语言模型的生成准确性与上下文相关性。然而，传统RAG方法仅依据查询-上下文相关性排序选择前几篇文档，常导致冗余和信息冲突。这一问题在无监督检索场景下尤为突出，因缺乏有效机制缓解，影响最终上下文选择效果。为此，我们提出一种完全无监督且无需训练的框架SMART（Selection using Matrices for Augmented Retrieval），用于问答任务中的上下文优化选择。SMART利用确定性点过程（DPPs）同时建模相关性、多样性与冲突，确保选出潜在高质量的上下文。跨多个数据集的实验表明，SMART显著提升了问答性能，超越了以往无监督上下文选择方法，展现出RAG中上下文选择的有前景策略。
原文摘要 · Abstract (English)
Retrieval-Augmented Generation (RAG) has greatly improved large language models (LLMs) by enabling them to generate accurate, contextually grounded responses through the integration of external information. However, conventional RAG approaches, which prioritize top-ranked documents based solely on query-context relevance, often introduce redundancy and conflicting information. This issue is particularly evident in unsupervised retrieval settings, where there are no mechanisms to effectively mitigate these problems, leading to suboptimal context selection. To address this, we propose Selection using Matrices for Augmented Retrieval (SMART) in question answering tasks, a fully unsupervised and training-free framework designed to optimize context selection in RAG. SMART leverages Determinantal Point Processes (DPPs) to simultaneously model relevance, diversity and conflict, ensuring the selection of potentially high-quality contexts. Experimental results across multiple datasets demonstrate that SMART significantly enhances QA performance and surpasses previous unsupervised context selection methods, showing a promising strategy for RAG.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。