---
id: 20260305-T0-17
title: "SuperLocalMemory：多智能体本地化记忆系统"
title_en: "SuperLocalMemory: Secure Multi-Agent Memory System"
url: https://ai.daily.yangsir.net/daily/20260305-T0-17
issue_date: 2026-03-05
publish_date: 2026-03-04T05:00:00.000Z
source_name: "arXiv cs.AI"
source_url: https://arxiv.org/abs/2603.02240
---

# SuperLocalMemory：多智能体本地化记忆系统

arXiv发布SuperLocalMemory系统，专为多智能体AI设计。该系统通过架构隔离和贝叶斯信任评分防御记忆投毒，同时通过自适应学习实现个性化检索。实验显示，其在OWASP ASI06攻击场景下错误率降低40%。开发者可将其用于敏感数据场景。

## English Version

**SuperLocalMemory: Secure Multi-Agent Memory System**

arXiv paper introduces SuperLocalMemory, a memory system for multi-agent AI using architectural isolation and Bayesian trust scoring to defend against memory poisoning. It achieves 40% lower error rates in OWASP ASI06 attack scenarios through adaptive learning, suitable for sensitive data applications.

---

**来源**：[arXiv cs.AI](https://arxiv.org/abs/2603.02240)

**详情页**：https://ai.daily.yangsir.net/daily/20260305-T0-17

---

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