Exploring Advanced LLM Multi-Agent Systems Based on Blackboard Architecture

TL;DR

This paper introduces a novel blackboard architecture for LLM multi-agent systems, enabling dynamic information sharing and consensus-building, leading to improved performance and efficiency in complex problem-solving tasks.

Contribution

It is the first to integrate blackboard architecture into LLM multi-agent systems, enhancing collaboration and adaptability in problem-solving.

Findings

Achieves state-of-the-art performance on commonsense, reasoning, and mathematical datasets.

Reduces token usage compared to existing static and dynamic MASs.

Demonstrates potential for complex, unstructured problem-solving.

Abstract

In this paper, we propose to incorporate the blackboard architecture into LLM multi-agent systems (MASs) so that (1) agents with various roles can share all the information and others' messages during the whole problem-solving process, (2) agents that will take actions are selected based on the current content of the blackboard, and (3) the selection and execution round is repeated until a consensus is reached on the blackboard. We develop the first implementation of this proposal and conduct experiments on commonsense knowledge, reasoning and mathematical datasets. The results show that our system can be competitive with the SOTA static and dynamic MASs by achieving the best average performance, and at the same time manage to spend less tokens. Our proposal has the potential to enable complex and dynamic problem-solving where well-defined structures or workflows are unavailable.