Harness-Zero: Harness Distillation via Agent-as-Harness· Harness-Zero: 通过代理系统蒸馏实现行为转移
Agent harnesses, the external systems that mediate model-environment interaction, can substantially improve agent performance, but their gains remain tied to the harness at deployment. Because the best harness varies across domains, instances, and models, a general-purpose agent must either settle for a suboptimal shared harness or route among an ever-growing set of specialized ones. We therefore study agent harness distillation: using a domain- or instance-optimized harness as training-time guidance and transferring the behaviors it induces into model weights, so that its gains survive under a single fixed target harness. The challenge is that the two harnesses differ in action space and available information, so guidance from the optimized harness cannot serve directly as supervision for
通过优化的代理系统指导训练，将特定领域的行为转移至通用模型中。
- 核心方法
- 研究了一种代理系统蒸馏方法，该方法通过在训练时使用特定领域或实例优化的代理系统作为指导，将由其诱导的行为转移到模型权重中，从而在使用单一固定目标代理系统时仍能保持性能提升。
- 适合谁读
- 研究者
- 要解决的问题
- 在部署时，优化的代理系统带来的性能提升无法保持，因为不同领域的最优代理系统各不相同。
- 关键实验
- 未提供
- 主要贡献
- 提出了一种新的代理系统蒸馏框架，使模型能够在不同环境下使用通用的代理系统保持高性能。
- 意义与局限
- 该研究为通用AI代理系统的发展提供了新的思路，但目前尚未经过充分实验验证。