让智能体像人一样复用经验流程，高效完成复杂网页任务。
Agent Workflow Memory
- 从过往任务中自动提取可复用的操作流程，动态引导智能体决策。
- 在两个主流网页导航数据集上，成功率提升24.6%至51.1%。
- 适用于在线和离线场景，跨任务/网站/领域泛化能力强。
尽管基于语言模型的智能体具备解决真实世界任务（如网页导航）的潜力，但当前方法在处理长周期、复杂动作轨迹的任务时仍存在困难。与人类能通过过往经验学习可复用的任务流程并指导未来行为不同，本文提出代理工作流记忆（Agent Workflow Memory, AWM），一种从训练样本或实时测试查询中提取常见重复操作流程的方法，并选择性地将这些流程提供给智能体以指导后续生成。AWM可灵活应用于离线与在线场景。在涵盖200多个领域、1000多个任务的Mind2Web和WebArena两大网页导航基准上进行实验，结果表明：相较于基线，AWM在Mind2Web和WebArena上的相对成功率分别提升24.6%和51.1%，同时显著减少成功解决WebArena任务所需的步数。此外，在线版AWM在跨任务、跨网站和跨领域评估中表现出稳健泛化能力，当训练-测试任务分布差异增大时，性能超越基线8.9至14.0个百分点。
原文摘要 · Abstract (English)
Despite the potential of language model-based agents to solve real-world tasks such as web navigation, current methods still struggle with long-horizon tasks with complex action trajectories. In contrast, humans can flexibly solve complex tasks by learning reusable task workflows from past experiences and using them to guide future actions. To build agents that can similarly benefit from this process, we introduce Agent Workflow Memory (AWM), a method for inducing commonly reused routines, i.e., workflows, and selectively providing workflows to the agent to guide subsequent generations. AWM flexibly applies to both offline and online scenarios, where agents induce workflows from training examples beforehand or from test queries on the fly. We experiment on two major web navigation benchmarks -- Mind2Web and WebArena -- that collectively cover 1000+ tasks from 200+ domains across travel, shopping, and social media, among others. AWM substantially improves the baseline results by 24.6% and 51.1% relative success rate on Mind2Web and WebArena while reducing the number of steps taken to solve WebArena tasks successfully. Furthermore, online AWM robustly generalizes in cross-task, website, and domain evaluations, surpassing baselines from 8.9 to 14.0 absolute points as train-test task distribution gaps widen.
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。