arXiv 2605.067022026-05-05cs.AIcs.CLcs.LG

CASCADE：基于案例的连续适应：在部署期间为大型语言模型进行持续适应

CASCADE: Case-Based Continual Adaptation for Large Language Models During Deployment

School of Artificial Intelligence, Jilin University（吉林大学人工智能学院）；Engineering Research Center of Knowledge-Driven Human-Machine Intelligence, Jilin University（吉林大学知识驱动人机智能工程研究中心）；International Center of Future Science, Jilin University（吉林大学未来科学国际中心）；Department of Informatics, King’s College London（伦敦国王学院信息学院）；The Alan Turing Institute（艾伦·图灵研究所）；AI Centre, Department of Computer Science, UCL（UCL计算机科学系人工智能中心）

Siyuan Guo, Yali Du, Hechang Chen, Yi Chang, Jun Wang

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出CASCADE框架，通过在部署期间持续学习提升LLM性能，实现20.9%的提升，并在多个领域任务中优于基线方法。

摘要：

大型语言模型（LLMs）已成为现代人工智能的核心基础，但其生命周期仍受训练与部署之间严格分割的限制，部署后学习效果显著下降。本文将部署时间学习（DTL）定义为LLM生命周期的第三阶段，使LLM代理在不修改模型参数的情况下通过经验提升自身。我们提出了CASCADE（CASe-based Continual Adaptation during DEployment），一种通用且原则性的框架，使LLM代理具备显式且持续演化的片段记忆。CASCADE将经验重用建模为上下文老虎机问题，使代理能够进行原则性的探索-利用权衡，并在长期交互中建立无遗憾保证。此设计使代理能够积累、选择和优化任务相关案例，将过去经验转化为可操作的知识。在16个多样化的任务中，包括医学诊断、法律分析、代码生成、网络搜索、工具使用和具身交互，CASCADE在零样本提示下将宏平均成功率提高了20.9%，并一致优于梯度和记忆基线方法。通过将部署重新定义为适应性学习过程，本文为持续改进人工智能系统奠定了基础。

英文摘要：

Large language models (LLMs) have become a central foundation of modern artificial intelligence, yet their lifecycle remains constrained by a rigid separation between training and deployment, after which learning effectively ceases. This limitation contrasts with natural intelligence, which continually adapts through interaction with its environment. In this paper, we formalise deployment-time learning (DTL) as the third stage in the LLM lifecycle that enables LLM agents to improve from experience during deployment without modifying model parameters. We present CASCADE (CASe-based Continual Adaptation during DEployment), a general and principled framework that equips LLM agents with an explicit, evolving episodic memory. CASCADE formulates experience reuse as a contextual bandit problem, enabling principled exploration-exploitation trade-offs and establishing no-regret guarantees over long-term interactions. This design allows agents to accumulate, select, and refine task-relevant cases, transforming past experience into actionable knowledge. Across 16 diverse tasks spanning medical diagnosis, legal analysis, code generation, web search, tool use, and embodied interaction, CASCADE improves macro-averaged success rate by 20.9% over zero-shot prompting while consistently outperforming gradient-based and memory-based baselines. By reframing deployment as an adaptive learning process, this work establishes a foundation for continually improving AI systems.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。