OMAC: A Broad Optimization Framework for LLM-Based Multi-Agent Collaboration

Shijun Li, Hilaf Hasson, Joydeep Ghosh

Keywords: Mutli-Agent Collaboration, Prompt Optimization, Agentic AI, Large Language Model

TL;DR: We introduce a holistic framework designed for the integral optimization of multi-agent collaboration with both agent functionality and collaboration structure.

Abstract:

Agents powered by advanced large language models (LLMs) have demonstrated impressive capabilities across diverse complex applications. Recently, Multi-Agent Systems (MAS), wherein multiple agents collaborate and communicate with each other, have exhibited enhanced capabilities in complex tasks, such as high-quality code generation and arithmetic reasoning. However, the development of such systems often relies on handcrafted methods, and the literature on systematic design and optimization of LLM-based MAS remains limited.

In this work, we introduce OMAC, a general framework designed for holistic optimization of LLM-based MAS. Specifically, we identify five key optimization dimensions for MAS, encompassing both agent functionality and collaboration structure. Building upon these dimensions, we first propose a general algorithm, utilizing two actors termed the Semantic Initializer and the Contrastive Comparator, to optimize any single dimension. Then, we present an algorithm for joint optimization across multiple dimensions. Extensive experiments demonstrate the superior performance of OMAC on diverse tasks against recent approaches.

Primary Area: applications to computer vision, audio, language, and other modalities

Submission Number: 2689