arXiv cs.CL· Zeyuan Zhao, Yuanyuan Shi, Peng Yu, Lingyue Fu, Shao Zhang, Ying Wen·· 12 小时前AI 评分38
OMCE：面向经验增强 LLM 智能体的测试时自适应元控制框架
Evolving Memory, Learning to Meta-Control: Test-Time Adaptation for Experience-augmented LLM Agents
AI 导读
OMCE 框架通过轻量级上下文赌博机元控制器，在测试时自适应中从任务反馈学习路由策略，在 DIRECT、EPISODE、GUIDE、DISTILL 四种认知动作间选择，决定是否调用经验、用哪种记忆格式以及何时整合新指导。
正文 · 原文
Abstract:Self-evolving agents accumulate experience from task streams, but existing frameworks typically rely on fixed pipelines to retrieve and utilize this experience, neglecting when and how to deploy it. We introduce Online Meta-Control of Experience (OMCE), a framework that updates experience memory and learns a meta-control policy from task feedback during test-time adaptation. OMCE employs a lightweight contextual-bandit meta-controller that uses compact retrieval signals to select among four cognitive actions: DIRECT, EPISODE, GUIDE, and DISTILL. This routing mechanism operationalizes three coupled decisions: whether to invoke experience, which memory format to utilize, and when to consolidate new guidance. OMCE evolves online, leveraging task feedback to update the memory bank and refine the routing policy via contextual bandit updates. Experiments on four online adaptation benchmarks across four backbones show that OMCE improves first-attempt success by up to 20.5 percentage points over the strongest compared baseline, while reducing per-task token use by a factor of 5.7 to 62.5 relative to ACE with GPT-5.2. Controlled ablations further support the benefits of learned routing and continued routing policy adaptation.
| Comments: | 23 pages, 5 figures. Submitted to ACL Rolling Review (October 2026) |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2611.01747 [cs.LG] |
| (or arXiv:2611.01747v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2611.01747 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Zeyuan Zhao [view email]
[v1]
Sun, 1 Nov 2026 09:47:28 UTC (10,013 KB)
来源：arXiv cs.CL · arxiv.org