---
id: 20261020-T0-02
title: "SAG2：不用嵌入向量的多跳检索，单个SQLite加BM25就能跑"
title_en: "SAG2 Delivers Embedding-Free Multi-Hop Retrieval on a Single SQLite Database"
url: https://ai.daily.yangsir.net/daily/20261020-T0-02
issue_date: 2026-10-20
publish_date: 2026-10-19T04:00:00.000Z
category: research
source_name: "arXiv cs.CL (NLP)"
source_url: https://arxiv.org/abs/2610.22588
---

# SAG2：不用嵌入向量的多跳检索，单个SQLite加BM25就能跑

新论文提出 SAG2，一种不依赖嵌入向量的多跳检索架构，核心组件只有一个 SQLite 数据库、BM25 和上下文证据机制。作者的主张是，强多跳检索不该以昂贵的语料预处理和复杂数据库基础设施为代价。相比主流向量检索方案，SAG2 省去了嵌入模型推理和向量库运维两个环节，部署门槛大幅降低。对需要搭建检索增强生成系统但缺少基础设施预算的团队，这类方案值得关注。

## English Version

**SAG2 Delivers Embedding-Free Multi-Hop Retrieval on a Single SQLite Database**

A new paper presents SAG2, an embedding-free multi-hop retrieval architecture built around a single SQLite database, BM25 and a contextual evidence mechanism. The authors argue that strong multi-hop retrieval should not require expensive corpus preparation or complex database infrastructure. Compared with mainstream vector retrieval, SAG2 eliminates both embedding model inference and vector database operations, sharply lowering the deployment barrier. The approach is worth attention from teams building retrieval-augmented generation without infrastructure budget.

---

**来源**：[arXiv cs.CL (NLP)](https://arxiv.org/abs/2610.22588)

**详情页**：https://ai.daily.yangsir.net/daily/20261020-T0-02

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*