衡量AI智能体自主性：探索一种结合代码审查的可扩展方法
Measuring AI agent autonomy: Towards a scalable approach with code inspection
AI总结：
针对现有AI智能体自主性评估依赖运行时评估、成本与风险较高的问题，提出一种基于代码审查的可扩展评估方法，可依据影响性与监督性分类体系为编排代码评分，并在AutoGen框架及应用中得到验证。
摘要：
AI智能体是能够自主完成复杂目标的AI系统，评估其自主性水平对于理解其潜在收益与风险至关重要。当前的自主性评估多聚焦于特定风险，且依赖运行时评估——即观察智能体运行过程中的行为。我们提出一种基于代码的自主性评估方法，无需运行AI智能体执行特定任务，从而降低运行时评估相关的成本与风险。借助该基于代码的框架，可依据一套评估自主性属性（影响性与监督性）的分类体系，对用于运行AI智能体的编排代码进行评分。我们在AutoGen框架及选定应用上验证了该方法。
英文摘要：
AI agents are AI systems that can achieve complex goals autonomously. Assessing the level of agent autonomy is crucial for understanding both their potential benefits and risks. Current assessments of autonomy often focus on specific risks and rely on run-time evaluations -- observations of agent actions during operation. We introduce a code-based assessment of autonomy that eliminates the need to run an AI agent to perform specific tasks, thereby reducing the costs and risks associated with run-time evaluations. Using this code-based framework, the orchestration code used to run an AI agent can be scored according to a taxonomy that assesses attributes of autonomy: impact and oversight. We demonstrate this approach with the AutoGen framework and select applications.
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。
