Mar, 2024
OPEx: 对体现教学指导中面向语言模型的组件式代理的分析
OPEx: A Component-Wise Analysis of LLM-Centric Agents in Embodied
Instruction Following
TL;DR在此研究中,提出了OPEx框架,它通过展示Observer、Planner和Executor的核心组成部分,深入分析了每个组件对Embodied Instruction Following任务性能的影响,并通过在TextWorld上部署多智能体对话策略来进一步提高性能。实验结果表明,引入大型语言模型和多智能体框架可以显著改善Embodied Instruction Following的结果。