BriefGPT.xyz
Feb, 2025
深度检索增强生成:为大型语言模型逐步思考检索
DeepRAG: Thinking to Retrieval Step by Step for Large Language Models
HTML
PDF
Xinyan Guan, Jiali Zeng, Fandong Meng, Chunlei Xin, Yaojie Lu...
TL;DR
本研究解决了大型语言模型在事实可信度上存在的严重幻觉问题,提出了一种名为DeepRAG的新框架,通过将检索增强推理建模为马尔可夫决策过程,优化了信息检索的效率和准确性。研究显示,DeepRAG在提高回答准确率21.99%的同时,显著提升了检索效率,具有很大的应用潜力。
Abstract
Large Language Models
(LLMs) have shown remarkable potential in reasoning while they still suffer from severe factual hallucinations due to timeliness, accuracy, and coverage of parametric knowledge. Meanwhile, integrating reasoning with
→