Agents centered around Large Language Models (LLMs) are now capable of automating mobile device operations for users. After fine-tuning to learn a user's mobile operations, these agents can adhere to high-level user instructions online. They execute tasks such as goal decomposition, sequencing of sub-goals, and interactive environmental exploration, until the final objective is achieved. However, privacy concerns related to personalized user data arise during mobile operations, requiring user confirmation. Moreover, users' real-world operations are exploratory, with action data being complex and redundant, posing challenges for agent learning. To address these issues, in our practical application, we have designed interactive tasks between agents and humans to identify sensitive information and align with personalized user needs. Additionally, we integrated Standard Operating Procedure (SOP) information within the model's in-context learning to enhance the agent's comprehension of complex task execution. Our approach is evaluated on the new device control benchmark AitW, which encompasses 30K unique instructions across multi-step tasks, including application operation, web searching, and web shopping. Experimental results show that the SOP-based agent achieves state-of-the-art performance without incurring additional inference costs, boasting an overall action success rate of 66.92%.

翻译：摘要: 以大型语言模型为核心的智能体现已能够为用户自动化移动设备操作。经微调学习用户操作后，这些智能体可在线遵循高级用户指令，执行目标分解、子目标排序及交互式环境探索等任务，直至最终目标达成。然而，移动操作中涉及个性化用户数据的隐私问题需要用户确认，且用户实际操作具有探索性，动作数据复杂冗余，给智能体学习带来挑战。为解决这些问题，我们在实际应用中设计了智能体与人类之间的交互任务，以识别敏感信息并与个性化用户需求对齐。同时，我们通过模型上下文学习集成了标准作业程序信息，以增强智能体对复杂任务执行的理解。本方法在新设备控制基准AitW上评估，该基准涵盖30K条跨多步骤任务的独特指令，包括应用操作、网页搜索及网络购物。实验结果表明，基于标准作业程序的智能体在不增加推理成本的情况下实现了最优性能，总体动作成功率达到66.92%。