Event ·Multi-source

EnvOpt 论文:优化智能体运行环境,Haiku 4.5 通过率提升 14.34%

一篇关于智能体环境优化的论文提出 EnvOpt 方法,目标对象是运行在 Slack、文档、工单等共享工作区中的智能体。它用一个编码智能体在无评分器、无参考答案的训练任务上运行目标智能体,编辑环境并根据验证集决定停止。在 AutomationBench 上,该方法让 Claude Haiku 4.5 的通过率提升 14.34%,成本降低 72.7%。同一环境迁移到 Opus 5 后成本下降 68.1%;叠加 Meta-Harness 的优化 harness,成本再降 83.4%,准确率增加 11.17%。

Current read

做智能体的朋友看看这篇:不改模型,只改智能体工作的环境,Haiku 4.5 通过率涨 14.34%,成本还能降七成。

8 sources47° AI heatLast updated 2026/11/6 14:27:01

Evidence

8 sources
Related source 1arXiv cs.AI
View source
Related source 2Perplexity
View source
Related source 4Yuchen Jin
View source
Related source 6eric zakariasson
View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.