推测性RAG：通过草稿生成增强检索增强生成
Speculative RAG: Enhancing Retrieval Augmented Generation through Drafting
AI总结：
推测性RAG利用大型通用LM验证小型专用LM并行生成的多个RAG草稿，通过分而治之减少令牌数并缓解位置偏差，在多个基准上实现最先进性能，准确率提升高达12.97%，延迟降低50.83%。
摘要：
检索增强生成（RAG）结合了大型语言模型（LLM）的生成能力与外部知识源，以提供更准确和最新的响应。近期的RAG进展侧重于通过迭代的LLM细化或通过额外指令调整获得的自我批判能力来改善检索结果。在这项工作中，我们引入了推测性RAG——一个利用较大的通用LM高效验证由较小的、经过蒸馏的专用LM并行生成的多个RAG草稿的框架。每个草稿由检索文档的不同子集生成，提供对证据的多角度视角，同时减少每个草稿的输入令牌数量。这种方法增强了对每个子集的理解，并减轻了长上下文中的潜在位置偏差。我们的方法通过将草稿生成委托给较小的专用LM来加速RAG，而较大的通用LM对草稿执行单次验证。大量实验表明，推测性RAG在TriviaQA、MuSiQue、PopQA、PubHealth和ARC-Challenge基准上实现了最先进的性能，同时降低了延迟。与传统的RAG系统相比，在PubHealth上，它显著提高了准确率高达12.97%，同时将延迟降低了50.83%。
英文摘要：
Retrieval augmented generation (RAG) combines the generative abilities of large language models (LLMs) with external knowledge sources to provide more accurate and up-to-date responses. Recent RAG advancements focus on improving retrieval outcomes through iterative LLM refinement or self-critique capabilities acquired through additional instruction tuning of LLMs. In this work, we introduce Speculative RAG - a framework that leverages a larger generalist LM to efficiently verify multiple RAG drafts produced in parallel by a smaller, distilled specialist LM. Each draft is generated from a distinct subset of retrieved documents, offering diverse perspectives on the evidence while reducing input token counts per draft. This approach enhances comprehension of each subset and mitigates potential position bias over long context. Our method accelerates RAG by delegating drafting to the smaller specialist LM, with the larger generalist LM performing a single verification pass over the drafts. Extensive experiments demonstrate that Speculative RAG achieves state-of-the-art performance with reduced latency on TriviaQA, MuSiQue, PopQA, PubHealth, and ARC-Challenge benchmarks. It notably enhances accuracy by up to 12.97% while reducing latency by 50.83% compared to conventional RAG systems on PubHealth.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
