Parallelized Planning-Acting for Efficient LLM-based Multi-Agent Systems

2025-03-05

AI Summary

Existing LLM-based multi-agent systems rely on a serial planning-execution paradigm, resulting in high response latency and poor adaptability in dynamic environments. To address this, we propose a “dual-thread planning-execution” architecture: (1) a novel interruptible parallel mechanism enabling concurrent planning and action execution; (2) a centralized shared memory module ensuring real-time state synchronization across agents; and (3) a skill library grounded in recursive task decomposition to autonomously drive the execution thread. Evaluated on complex Minecraft tasks, our approach achieves a 72% planning-execution overlap rate and improves task completion efficiency by 2.3× compared to baseline methods. It significantly enhances system responsiveness and environmental adaptability. This work establishes a new low-latency decision-making paradigm for LLM-powered multi-agent systems.

Technology Category

• Multiagent Systems: Multiagent Planning
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Recent advancements in Large Language Model(LLM)-based Multi-Agent Systems(MAS) have demonstrated remarkable potential for tackling complex decision-making tasks. However, existing frameworks inevitably rely on serialized execution paradigms, where agents must complete sequential LLM planning before taking action. This fundamental constraint severely limits real-time responsiveness and adaptation, which is crucial in dynamic environments with ever-changing scenarios. In this paper, we propose a novel parallelized planning-acting framework for LLM-based MAS, featuring a dual-thread architecture with interruptible execution to enable concurrent planning and acting. Specifically, our framework comprises two core threads:(1) a planning thread driven by a centralized memory system, maintaining synchronization of environmental states and agent communication to support dynamic decision-making; and (2) an acting thread equipped with a comprehensive skill library, enabling automated task execution through recursive decomposition. Extensive experiments on challenging Minecraft demonstrate the effectiveness of the proposed framework.

Problem

• Enhance real-time responsiveness in dynamic environments
• Enable concurrent planning and acting in multi-agent systems
• Overcome limitations of serialized execution in LLM-based systems

Innovation

• Parallelized planning-acting framework for LLM-based MAS
• Dual-thread architecture with interruptible execution
• Centralized memory system and comprehensive skill library