---
id: 20260319-T0-19
title: "记忆代理路由：降低成本并减少干扰"
title_en: "Cost-Sensitive Store Routing for Memory Agents"
url: https://ai.daily.yangsir.net/daily/20260319-T0-19
issue_date: 2026-03-19
publish_date: 2026-03-18T04:00:00.000Z
category: research
source_name: "arXiv cs.AI"
source_url: https://arxiv.org/abs/2603.15658
---

# 记忆代理路由：降低成本并减少干扰

加州大学伯克利分校提出'成本敏感存储路由'方法，优化记忆代理的检索效率。现有系统对所有存储进行查询，增加成本和无关信息。新方法智能选择最相关的存储，计算成本降低60%，准确率提升28%。开发者可用此构建更高效的记忆增强AI系统，如智能客服或个性化推荐。

## English Version

**Cost-Sensitive Store Routing for Memory Agents**

UC Berkeley researchers propose 'cost-sensitive store routing' to optimize memory agent retrieval efficiency. Existing systems query all stores, increasing costs and irrelevant information. The new method intelligently selects most relevant stores, reducing computation costs by 60% while improving accuracy by 28%. Developers can use this to build more efficient memory-enhanced AI systems like intelligent customer service or personalized recommendations.

---

**来源**：[arXiv cs.AI](https://arxiv.org/abs/2603.15658)

**详情页**：https://ai.daily.yangsir.net/daily/20260319-T0-19

---

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