The Geometry of Dialogue: Graphing Language Models to Reveal Synergistic Teams for Multi-Agent Collaboration

TL;DR

This paper introduces a graph-based framework for automatically forming synergistic teams of language models for multi-agent collaboration, without requiring prior knowledge of models, leading to improved performance on benchmarks.

Contribution

It proposes an interaction-centric, graph-based method for automatic team formation of LLMs that reveals latent specializations and enhances collaborative performance.

Findings

Discovered functionally coherent model groups reflecting their specializations

Teams identified by the method outperform random baselines on benchmarks

Achieved accuracy comparable to manually curated teams

Abstract

While a multi-agent approach based on large language models (LLMs) represents a promising strategy to surpass the capabilities of single models, its success is critically dependent on synergistic team composition. However, forming optimal teams is a significant challenge, as the inherent opacity of most models obscures the internal characteristics necessary for effective collaboration. In this paper, we propose an interaction-centric framework for automatic team composition that does not require any prior knowledge including their internal architectures, training data, or task performances. Our method constructs a "language model graph" that maps relationships between models from the semantic coherence of pairwise conversations, and then applies community detection to identify synergistic model clusters. Our experiments with diverse LLMs demonstrate that the proposed method discovers functionally coherent groups that reflect their latent specializations. Priming conversations with specific topics identified synergistic teams which outperform random baselines on downstream benchmarks and achieve comparable accuracy to that of manually-curated teams based on known model specializations. Our findings provide a new basis for the automated design of collaborative multi-agent LLM teams.