---
id: 20260709-T0-01
title: "新论文：将记忆放入LLM循环内部，提升智能体工作记忆"
title_en: "Paper: In-Process Retrieval as Extended Working Memory for Language Agents"
url: https://ai.daily.yangsir.net/daily/20260709-T0-01
issue_date: 2026-07-09
publish_date: 2026-07-08T04:00:00.000Z
category: research
source_name: "arXiv cs.AI"
source_url: https://arxiv.org/abs/2607.05690
---

# 新论文：将记忆放入LLM循环内部，提升智能体工作记忆

一篇 arXiv 新论文提出“内存循环”（Memory in the Loop）架构，将知识检索从每轮一次的外部查询，改为每步都读写的内存循环。传统智能体每轮只能外部查询一次记忆，新方法将记忆移入推理循环内部，每步均可读写。实验证明，该架构显著提升智能体在长序列任务中的表现，尤其需要持续引用上下文的场景。

## English Version

**Paper: In-Process Retrieval as Extended Working Memory for Language Agents**

A new arXiv paper proposes 'Memory in the Loop', which moves knowledge retrieval inside the agent's reasoning loop. Instead of querying external memory once per turn, the agent reads and writes memory on every step. Experiments show this architecture significantly improves performance on long-horizon tasks requiring continuous context reference.

---

**来源**：[arXiv cs.AI](https://arxiv.org/abs/2607.05690)

**详情页**：https://ai.daily.yangsir.net/daily/20260709-T0-01

---

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