Towards Multi-Agent Reasoning Systems for Collaborative Expertise Delegation: An Exploratory Design Study

TL;DR

This paper explores how to design multi-agent reasoning systems for collaboration, analyzing key factors like expertise alignment, collaboration methods, and system scale to improve collective reasoning performance.

Contribution

It systematically investigates the effects of expertise alignment, collaboration paradigms, and system scale on multi-agent reasoning, providing practical guidelines and identifying key trade-offs.

Findings

Expertise alignment benefits are domain-specific.

Diversity-driven collaboration outperforms structured workflows.

Scaling with expertise improves reasoning but introduces communication trade-offs.

Abstract

Designing effective collaboration structure for multi-agent LLM systems to enhance collective reasoning is crucial yet remains under-explored. In this paper, we systematically investigate how collaborative reasoning performance is affected by three key design dimensions: (1) Expertise-Domain Alignment, (2) Collaboration Paradigm (structured workflow vs. diversity-driven integration), and (3) System Scale. Our findings reveal that expertise alignment benefits are highly domain-contingent, proving most effective for contextual reasoning tasks. Furthermore, collaboration focused on integrating diverse knowledge consistently outperforms rigid task decomposition. Finally, we empirically explore the impact of scaling the multi-agent system with expertise specialization and study the computational trade off, highlighting the need for more efficient communication protocol design. This work provides concrete guidelines for configuring specialized multi-agent system and identifies critical architectural trade-offs and bottlenecks for scalable multi-agent reasoning. The code will be made available upon acceptance.