让弱模型通过专家修正局部错误，实现低成本高效协作。
Co-Evolving Harnesses and Models: On-Policy Correction Helps Weaker Models Catch Up Where Imitation Fails
- 用专家纠正弱模型在自执行中的单步错误，而非全轨迹模仿。
- 在7个企业任务中，性能提升4至30点，避免模仿导致的退化。
- 适合需要低成本部署的领域专用智能体场景。
智能体赋能系统（即模型提示、工具集、执行钩子和上下文管理结构）是决定智能体任务成功率的关键因素。自动化赋能系统演化可使小型模型以远低于前沿模型的成本，在特定领域任务上表现良好。由于赋能系统与模型权重共同决定行为，我们探究如何结合赋能系统演化与轻量微调。在七个企业级智能体任务中，先用弱模型演化出赋能系统，再发现更强专家能更有效地使用该系统，暗示专家监督可弥补剩余差距。然而，将弱模型在专家完整轨迹上的训练反而导致所有七项任务性能下降4至30分（基于Qwen3-Coder和Gemma 4），尽管同一方法在未演化的赋能系统下有效。分析表明，模仿虽传递知识并增加对系统组件的使用，但破坏了模型与赋能系统的适配性：弱模型采纳专家的规划策略，却缺乏执行能力，不再匹配为其原生规划风格所设计的赋能系统。为此，我们提出一种基于元级最大似然估计（MLE）代理的在线策略专家修正管道，自动定位弱模型自执行中的失败步骤，并仅请求专家重写该步骤。此方法保留了模型的规划风格，同时融合赋能系统演化与模型适应的优势。结果揭示并解决了赋能系统与权重更新间的冲突，为领域特定企业任务提供了一种兼容性保持的经济型协同演化方案。
原文摘要 · Abstract (English)
Agent harnesses (the system prompt, tool set, execution hooks, and context-management scaffolding around a model) are a critical determinant of agentic task success. Automated harness evolution can enable smaller models to perform well on domain-specific tasks at a fraction of frontier-model cost. Since both the harness and model weights shape behavior, we ask how harness evolution and lightweight fine-tuning should be combined. Across seven enterprise agent tasks, we first evolve a harness with the weaker model, then find that a stronger expert often uses it more effectively, suggesting expert supervision could close the remaining gap. However, training the weaker model on the expert's complete trajectories under the evolved harness backfires: performance regresses on all seven tasks by 4 to 30 points across Qwen3-Coder and Gemma 4, even though the same procedure helps under the unevolved harness. Our analysis shows that imitation transfers knowledge and increases scaffold usage, but disrupts model-harness fit: the weaker model adopts the expert's planning strategy without the competence to execute it and no longer matches the harness evolved around its native planning style. We therefore develop an on-policy expert-correction pipeline, automated by a meta-level MLE agent, that localizes the failing turn in the weaker model's own rollout and asks the expert to rewrite only that turn. This preserves the model's planning style and combines the gains of harness evolution and model adaptation. Our results identify and resolve a source of contention between harness and weight updates, yielding a compatibility-preserving recipe for economical co-evolution on domain-specific enterprise tasks.
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。