这篇论文测试了智能体能否训练其他智能体，发现它们在第一步就锁定了策略，后续改进效果有限。作者尝试了三种方法，但智能体仍缺乏在执行中重新考虑策略的能力。
该论文分析了大量公开的智能体后训练轨迹。研究发现，智能体在第一步就锁定了训练策略，后续仅进行局部调整。论文尝试了三种改进方法：经验驱动的脚手架提升了GSM8K和HumanEval的分数，但策略仍保持冻结。人类指导能改变初始选择，但训练开始后智能体又回到局部循环。额外推理计算对简单任务有效，对最难任务几乎无效。
Finally a good paper testing whether agents can really post-train other agents. (bookmark it) They...
Finally a good paper testing whether agents can really post-train other agents. (bookmark it) They analyzed a large corpus of publicly released post-training trajectories. Across tasks, the agent locks in its training strategy at the very first step and spends the entire remaining budget on local adjustments inside it. They then tried three escalating fixes. An experience-driven scaffold lifted execution broadly, worth 12.6 points on GSM8K and 40.8 on HumanEval, and the strategy stayed frozen. Human guidance redirected the opening choice, and the agent slid back into local loops once training began. Extra inference compute paid off on easy tasks and did almost nothing on the hardest one. What agents lack here is a way to reconsider strategy while execution is still running. Paper: arxiv.org/abs/2608.19072 Track more trending AI papers in our academy: academy.dair.ai 💬 0 🔄 1 ❤️ 7 👀 963 📊 2 ⚡
- arXiv cs.LG08-19 16:17原文