MSI-Agent: Incorporating Multi-Scale Insight into Embodied Agents for Superior Planning and Decision-Making

TL;DR

MSI-Agent enhances embodied agents' planning and decision-making by effectively summarizing and utilizing multi-scale insights, leading to improved performance and robustness in domain-shifting scenarios.

Contribution

This paper introduces MSI-Agent, a novel multi-scale insight framework that improves LLM-based agents' decision-making through a structured insight generation and selection pipeline.

Findings

MSI-Agent outperforms other insight strategies in GPT-3.5 planning tasks.

MSI exhibits better robustness under domain shifts.

Effective multi-scale insight management enhances long-term decision-making.

Abstract

Long-term memory is significant for agents, in which insights play a crucial role. However, the emergence of irrelevant insight and the lack of general insight can greatly undermine the effectiveness of insight. To solve this problem, in this paper, we introduce Multi-Scale Insight Agent (MSI-Agent), an embodied agent designed to improve LLMs' planning and decision-making ability by summarizing and utilizing insight effectively across different scales. MSI achieves this through the experience selector, insight generator, and insight selector. Leveraging a three-part pipeline, MSI can generate task-specific and high-level insight, store it in a database, and then use relevant insight from it to aid in decision-making. Our experiments show that MSI outperforms another insight strategy when planning by GPT3.5. Moreover, We delve into the strategies for selecting seed experience and insight, aiming to provide LLM with more useful and relevant insight for better decision-making. Our observations also indicate that MSI exhibits better robustness when facing domain-shifting scenarios.