MEMTIER通过分层记忆架构显著提升长时自主AIAgent的记性与推理能力。
MEMTIER: Tiered Memory Architecture and Retrieval Bottleneck Analysis for Long-Running Autonomous AI Agents
- 设计三层次记忆结构，引入加权检索与认知权重更新机制
- 在500题长时记忆测试中准确率提升33个百分点至38.2%
- 可在6GB消费级显卡上本地运行，适合轻量级长期智能体
长时间运行的自主AI代理面临严重的记忆一致性问题：在72小时运行窗口内，工具执行成功率因现有扁平文件记忆系统的四种叠加失效模式下降14个百分点。我们提出MEMTIER，一种为OpenClaw代理运行时设计的三层记忆架构，包含结构化的事件记忆JSONL存储、五信号加权检索引擎、注意力归因的认知权重更新环路、异步合并守护进程将事件事实提升至语义层级，以及基于PPO的检索权重自适应策略（基础设施已验证；性能提升待最终版本）。在完整的500题LongMemEval-S基准测试（Wu et al., 2025）中，MEMTIER在消费级6GB GPU上使用Qwen2.5-7B模型实现准确率0.382、F1值0.412，相比全上下文基线（0.050 → 0.382，即5% → 38%）提升33个百分点。使用DeepSeek-V4-Flash进行事实预填充后，单会话召回率达0.686–0.714，超过论文中RAG BM25 GPT-4o基线（0.560）。时间推理能力提升至0.323，多会话合成达0.173，表明结构化语义预填充质变地提升了轻量检索的能力。所有阶段均可在配备6GB GPU的消费级笔记本上本地运行。
原文摘要 · Abstract (English)
Long-running autonomous AI agents suffer from a well-documented memory coherence problem: tool-execution success rates degrade 14 percentage points over 72-hour operation windows due to four compounding failure modes in existing flat-file memory systems. We present MEMTIER, a tripartite memory architecture for the OpenClaw agent runtime that introduces a structured episodic JSONL store, a five-signal weighted retrieval engine, an attention-attributed cognitive weight update loop, an asynchronous consolidation daemon promoting episodic facts to a semantic tier, and a PPO-based policy framework for adapting retrieval weights (infrastructure validated; performance gains pending camera-ready). On the full 500-question LongMemEval-S benchmark (Wu et al., 2025), MEMTIER achieves Acc=0.382, F1=0.412 with Qwen2.5-7B on a consumer 6GB GPU - a +33 percentage point improvement over the full-context baseline (0.050 -> 0.382, i.e., 5% -> 38%). With DeepSeek-V4-Flash fact pre-population, single-session recall reaches 0.686-0.714, exceeding the paper's RAG BM25 GPT-4o baseline (0.560) on those categories. Temporal reasoning rises to 0.323 and multi-session synthesis to 0.173, demonstrating that structured semantic pre-population qualitatively changes what lightweight retrieval can achieve. All phases run locally on a consumer laptop with a 6GB GPU.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。