---
id: 20260806-T0-05
title: "MemArena：首个面向端侧智能体的个人记忆基准测试，可处理上万条对话"
title_en: "MemArena: New On-Device Benchmark for AI Memory Assistants"
url: https://ai.daily.yangsir.net/daily/20260806-T0-05
issue_date: 2026-08-06
publish_date: 2026-08-05T04:00:00.000Z
category: research
source_name: "arXiv cs.CL (NLP)"
source_url: https://arxiv.org/abs/2608.02613
---

# MemArena：首个面向端侧智能体的个人记忆基准测试，可处理上万条对话

MemArena 是一个新的基准测试框架，专门用于评估端侧部署的个人记忆助手。该基准强调三个核心维度：高密度活动交互、自我中心视角和连贯性。测试规模较大，旨在模拟真实设备上的复杂私人对话场景，帮助开发者衡量开源模型在本地处理个人记忆任务时的表现。

## English Version

**MemArena: New On-Device Benchmark for AI Memory Assistants**

MemArena is a new benchmark designed for evaluating on-device personal memory assistants. It focuses on three key aspects: activity-dense interaction, ego-centric perspective, and coherence. The benchmark tests models on large-scale simulated private conversations, helping developers assess how well open-weight models handle personal memory tasks locally.

---

**来源**：[arXiv cs.CL (NLP)](https://arxiv.org/abs/2608.02613)

**详情页**：https://ai.daily.yangsir.net/daily/20260806-T0-05

---

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