研究发现弱行动模型也能充当有效评判者和教师
AI 总结
新论文《Actor, Judge, or Teacher? Understanding Decision Models in Long-Horizon Agents》研究行动能力较弱的语言模型,能否有效评估动作或教授长程智能体。作者报告称,能够查看完整轨迹的评判者可区分有用步骤,JEV教师则让采用GRPO训练的Qwen2.5-1.5B策略在ALFWorld上的成功率从72.8%升至82.9%,在WebShop上从56.8%升至71.4%。这些结果由作者自行报告,尚未得到独立验证。
深度见解
该论文提供了一个具体研究结果:语言模型在长程智能体系统中的价值可能不止于直接选择动作,并在两个强化学习基准中报告了策略性能提升。
Computer Science > Artificial Intelligence
arXiv:2611.17167 (cs)
[Submitted on 12 Nov 2026]
Title:Actor, Judge, or Teacher? Understanding Decision Models in Long-Horizon Agents
Authors:Haochen Liu, Qianhao Zhu, Zixuan Wang, Dingming Li, Tongbo Chen, Peng Qiu, Yingqi Qu, Weiming Lu, Jun Xiao, Yongliang Shen
View a PDF of the paper titled Actor, Judge, or Teacher? Understanding Decision Models in Long-Horizon Agents, by Haochen Liu and 9 other authors