让安卓智能体自动造工具，效率提升20%，准确率高4.47%。
Improving Proficiency and Efficiency of Android GUI Agents via Self-Generating Tool Actions
- 用自动生成的Python工具函数扩展智能体能力，减少人工开发成本。
- 在三个基准上平均性能提升4.47个百分点，交互次数减少20.05%。
- 通过关联测试提升覆盖度，适合做安卓自动化测试的研究者使用。
采用混合动作空间（包含GUI动作和工具动作，如通过API访问应用数据）的安卓智能体研究仍不充分，主要因工具构建成本过高。为此，我们提出DroidTool框架，使智能体能自动生成工具（以操作应用状态的Python函数实现）。该框架采用代理式工作流，包含提议、实现、测试生成与执行、修复四个阶段。特别地，在验证工具时，通过跨工具关联测试构建合适前置条件，提升测试覆盖率，而非单独测试每个工具。在AndroidWorld、B-MoCA和MobileSafetyBench三个代表性基准上，增强工具后的智能体平均性能提升约4.47个百分点，交互次数减少约20.05%。
原文摘要 · Abstract (English)
Android agents using a hybrid action space that combines GUI actions and tool actions (e.g., accessing application data via APIs) remain largely underexplored, mainly due to the excessive effort required to create tools. To address this gap, we introduce DroidTool, a framework for augmenting the agents with self-generated tools, which are realized as Python functions operating on application states (e.g., a database). To create tools with minimal human labor, DroidTool employs an agentic workflow featuring stages: proposal, implementation, test generation and execution, and repair. Notably, when testing the created tools for verification, it constructs relational tests across relevant tools for natural preparation of appropriate test preconditions and improved test coverage, rather than testing each tool separately. The GUI agents augmented with the generated tools achieved approximately 4.47%p higher performance with approximately 20.05% fewer interactions than the GUI-only agents, averaged across representative benchmarks: AndroidWorld, B-MoCA, and MobileSafetyBench.
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。