Google 研究：Harness-Zero 在训练后移除专用 harness，任务成功率反升
Google 这篇论文把 agent harness 蒸馏进模型，移除专用 harness 后成功率反而从 23.3% 涨到 44.3%，比带 harness 还高，做 agent 的都该看看。
Google 与合作者的论文提出 Harness-Zero，研究能否把 agent harness 蒸馏进模型本身。
移除专用 harness 后，宏观任务成功率从 23.3% 升至 44.
3%，高于基础模型带 harness 时的 41.7%。
方法是在训练时使用优化过的 harness，由 harnessing agent 在部署动作空间中纠正学生模型的响应，纠正后的运行轨迹作为训练演示。在知识工作、工具使用和科学领域的 28 种 harness 诱导行为上，平均恢复 82.
3%。对于使用相同进化 harness 的前沿模型，agent-as-harness 形式也优于 code-as-harness 形式。
Super interesting paper from Google and colleagues. It studies where it's possible to distill an agent harness. With the specialized harness removed, macro task success goes from 23.3% to 44.3%. That is higher than the 41.7% the base model reaches with the harness attached. Harness-Zero uses the optimized harness only during training. The optimized harness and the deployment harness have different action spaces, so a harnessing agent guided by the optimized harness corrects the student's responses in the deployment action space before they run. Those corrected runs become the training demonstrations. Across 28 harness-induced behaviors in knowledge work, tool use and science, 82.3% are recovered on average. For frontier models using the same evolved harness, the agent-as-harness form also beats the code-as-harness form. It remains to be seen how robust the approach is, but it's very interesting to see potential in harness distillation. Paper: arxiv.org/abs/2609.24974 Chat with Paper: academy.dair.ai/papers/harness…📊 10 ⚡
本条情报归属 AITOP「论文」核心栏目，重点聚焦北美前沿实验室与硅谷科技巨头。在跨区域产业观察坐标中，该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线，反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括：Harness-Zero、Google、agent、蒸馏，读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明：本文原始线索采编自「elvis」（@omarsar0），经 AITOP 多维评分管线结构化提炼。
