用大模型生成的有用性信号训练检索模型，提升精度且速度超快。
Aligning Dense Retrievers with LLM Utility via Distillation
- 通过模仿大模型的困惑度降低分布，训练双编码器直接注入有用性信号。
- 在QASPER上召回率提升30.59%，速度比高效重排序方法快180倍以上。
- 适合需要高精度与低延迟的实时问答系统使用。
密集向量检索是检索增强生成（RAG）的实际核心，但相似性搜索常受限于精度。相反，基于大模型重排序的效用方法虽性能优越，却计算开销大且易受困惑度估计噪声影响。本文提出实用对齐嵌入（UAE），将两者优势融合为高效高精度的检索方法。将检索建模为分布匹配问题，使用效用调制InfoNCE目标训练双编码器，模仿由困惑度降低导出的效用分布，直接在嵌入空间注入分级效用信号，无需测试时进行大模型推理。在QASPER基准上，UAE相较于强基线BGE-Base，召回率@1提升30.59%，MAP提升30.16%，Token F1提升17.3%。关键的是，UAE速度超过180倍于高效大模型重排序方法，同时保持竞争力，证明对齐生成效用可实现大规模可靠上下文生成。
原文摘要 · Abstract (English)
Dense vector retrieval is the practical backbone of Retrieval- Augmented Generation (RAG), but similarity search can suffer from precision limitations. Conversely, utility-based approaches leveraging LLM re-ranking often achieve superior performance but are computationally prohibitive and prone to noise inherent in perplexity estimation. We propose Utility-Aligned Embeddings (UAE), a framework designed to merge these advantages into a practical, high-performance retrieval method. We formulate retrieval as a distribution matching problem, training a bi-encoder to imitate a utility distribution derived from perplexity reduction using a Utility-Modulated InfoNCE objective. This approach injects graded utility signals directly into the embedding space without requiring test-time LLM inference. On the QASPER benchmark, UAE improves retrieval Recall@1 by 30.59%, MAP by 30.16% and Token F1 by 17.3% over the strong semantic baseline BGE-Base. Crucially, UAE is over 180x faster than the efficient LLM re-ranking methods preserving competitive performance, demonstrating that aligning retrieval with generative utility yields reliable contexts at scale.
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。