Tag: Semantic Understanding
All the articles with the tag "Semantic Understanding".
-
Multiple Weaks Win Single Strong: Large Language Models Ensemble Weak Reinforcement Learning Agents into a Supreme One
本文提出LLM-Ens框架,利用大型语言模型(LLMs)通过语义状态分类和动态代理选择增强强化学习模型集成,在Atari基准上显著提升性能,最高较基线方法提升51.2%。