Position: Behavioral Systems Require Behavioral Tests

Position: Behavioral Systems Require Behavioral Tests

立场：行为系统需要行为测试

Abstract: Artificial agentic systems increasingly operate as behavioral systems by interacting with dynamic environments, pursuing goals, and adapting over time. Yet, current evaluation methods largely focus on performance outcomes, not the underlying behavioral processes that produce them. 摘要： 人工智能代理系统正日益作为行为系统运行，通过与动态环境交互、追求目标并随时间进行自我调整。然而，当前的评估方法主要关注性能结果，而非产生这些结果的底层行为过程。

This paper argues that AI agents must be evaluated like other behavioral systems: through systematic observation, perturbation, and interpretation of their actions. We draw on lessons from the behavioral sciences to motivate this position, and propose a research agenda focused on developing rigorous behavioral tests. 本文认为，必须像评估其他行为系统一样评估人工智能代理：通过对其行为进行系统性的观察、扰动和解读。我们借鉴行为科学的经验来阐述这一立场，并提出了一项专注于开发严谨行为测试的研究议程。

These include methods for recovering decision strategies from action sequences, constructing environments that isolate behavioral differences, and probing emergent dynamics in multi-agent systems. Taken together, these directions offer a roadmap for developing a science of AI behavior. 这些测试包括：从动作序列中恢复决策策略的方法、构建能够隔离行为差异的环境，以及探测多智能体系统中的涌现动力学。总而言之，这些方向为发展人工智能行为科学提供了一份路线图。