MSI-Agent: Incorporating Multi-Scale Insight into Embodied Agents for Superior Planning and Decision-Making
Published in EMNLP 2024, 2024
本研究提出了MSI-Agent，一个创新的具身智能体框架，通过整合多尺度洞察来增强智能体的规划和决策能力。该方法能够从不同层面理解和分析任务环境，从而做出更优的决策。
Recommended citation: Fu, D., Qi, B., Gao, Y., Jiang, C., Dong, G., & Zhou, B. (2024). MSI-Agent: Incorporating Multi-Scale Insight into Embodied Agents for Superior Planning and Decision-Making. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing.
Download Paper