arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2508.08053cs.AI
收藏

AdaptFlow:基于元学习的自适应工作流优化

AdaptFlow: Adaptive Workflow Optimization via Meta-Learning

Peking University(北京大学);University of Chinese Academy of Sciences(中国科学院大学);Microsoft(微软公司);Shanghai Jiaotong University(上海交通大学)

Runchuan Zhu, Bowen Jiang, Lingrui Mei, Fangkai Yang, Lu Wang, Haoxiang Gao, Fengshuo Bai, Pu Zhao, Qingwei Lin, Saravan Rajmohan, Dongmei Zhang

AI总结:

针对现有智能体工作流依赖静态模板导致适应性差的问题,提出基于元学习的框架AdaptFlow,通过双层优化方案学习可泛化的工作流初始化并实现快速子任务级适应,在多任务基准上取得最优结果。

摘要:

大型语言模型(LLM)的最新进展激发了对智能体工作流的日益关注,这些工作流是旨在解决复杂任务的LLM调用的结构化序列。然而,现有方法通常依赖于静态模板或手动设计的工作流,这限制了对多样化任务的适应性并阻碍了可扩展性。我们提出了AdaptFlow,这是一个受模型无关元学习(MAML)启发的基于自然语言的元学习框架。AdaptFlow学习一个可泛化的工作流初始化,从而实现快速的子任务级适应。它采用双层优化方案:内层循环利用LLM生成的反馈针对特定子任务细化工作流,而外层循环更新共享初始化以在跨任务中表现良好。这种设置允许AdaptFlow通过语言引导的修改来适应初始化的工作流,从而有效地泛化到未见任务。在问答、代码生成和数学推理基准上进行评估,AdaptFlow始终优于手动构建和自动搜索的基线,取得了最先进的结果,并在跨任务和模型中表现出强大的泛化能力。源代码和数据可在https://github.com/microsoft/DKI_LLM/tree/AdaptFlow/AdaptFlow获取。

英文摘要:

Recent advances in large language models (LLMs) have sparked growing interest in agentic workflows, which are structured sequences of LLM invocations intended to solve complex tasks. However, existing approaches often rely on static templates or manually designed workflows, which limit adaptability to diverse tasks and hinder scalability. We propose AdaptFlow, a natural language-based meta-learning framework inspired by model-agnostic meta-learning (MAML). AdaptFlow learns a generalizable workflow initialization that enables rapid subtask-level adaptation. It employs a bi-level optimization scheme: the inner loop refines the workflow for a specific subtask using LLM-generated feedback, while the outer loop updates the shared initialization to perform well across tasks. This setup allows AdaptFlow to generalize effectively to unseen tasks by adapting the initialized workflow through language-guided modifications. Evaluated across question answering, code generation, and mathematical reasoning benchmarks, AdaptFlow consistently outperforms both manually crafted and automatically searched baselines, achieving state-of-the-art results with strong generalization across tasks and models. The source code and data are available at https://github.com/microsoft/DKI_LLM/tree/AdaptFlow/AdaptFlow.

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。

↑