优化小模型RAG的提示词设计，显著提升多跳问答准确率。
Evaluating Prompt Engineering Techniques for RAG in Small Language Models: A Multi-Hop QA Approach
- 测试24种提示模板，融合文献方法与新型混合设计。
- 在HotpotQA上最高提升84.5%，较标准提示提升6%。
- 为资源受限环境下的小模型RAG提供实用优化方案。
检索增强生成（RAG）通过整合外部知识提升语言模型的事实准确性。尽管在大模型中已有广泛研究，针对小语言模型（SLMs）的RAG优化仍存在关键空白，尤其在需要复杂推理的多跳问答任务中。提示模板设计是影响性能的关键因素，但尚未充分探索。本文开展大规模实证研究，评估24种不同提示模板在HotpotQA数据集上的表现，涵盖标准RAG提示、文献中9种优质技术及14种新型混合变体，均在Qwen2.5-3B Instruct和Gemma3-4B-It两个主流SLM上测试。基于18720个测试样本，结果显示：在Qwen2.5上最高提升83%，在Gemma3-4B-It上最高提升84.5%，相较标准提示分别提升6%。研究还提供了可操作的提示设计分析与建议，助力资源受限环境下SLM-RAG系统的实际部署。
原文摘要 · Abstract (English)
Retrieval Augmented Generation (RAG) is a powerful approach for enhancing the factual grounding of language models by integrating external knowledge. While widely studied for large language models, the optimization of RAG for Small Language Models (SLMs) remains a critical research gap, particularly in complex, multi-hop question-answering tasks that require sophisticated reasoning. In these systems, prompt template design is a crucial yet under-explored factor influencing performance. This paper presents a large-scale empirical study to investigate this factor, evaluating 24 different prompt templates on the HotpotQA dataset. The set includes a standard RAG prompt, nine well-formed techniques from the literature, and 14 novel hybrid variants, all tested on two prominent SLMs: Qwen2.5-3B Instruct and Gemma3-4B-It. Our findings, based on a test set of 18720 instances, reveal significant performance gains of up to 83% on Qwen2.5 and 84.5% on Gemma3-4B-It, yielding an improvement of up to 6% for both models compared to the Standard RAG prompt. This research also offers concrete analysis and actionable recommendations for designing effective and efficient prompts for SLM-based RAG systems, practically for deployment in resource-constrained environments.
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。