Search papers, labs, and topics across Lattice.
This paper analyzes multi-agent LLM deliberation using the Friedkin-Johnsen (FJ) opinion dynamics model, demonstrating that the model's parameters are input-dependent, effectively creating a mixture of experts. They show that this dynamic routing based on competence allows multi-agent systems to surpass single agents and static ensembles. The paper then explores how influence emerges through observable proxies like self-assessed confidence, perceived confidence, and initial alignment.
Multi-agent LLM systems beat single agents not just through diverse opinions, but by dynamically weighting agents based on input-specific competence, turning deliberation into a learned routing mechanism.
The effectiveness of multi-agent LLM deliberation depends not only on the agents' individual predictions, but also on how they communicate and collaborate. We study this mechanism through the lens of Friedkin-Johnsen (FJ) opinion dynamics, a tractable model for analyzing stubbornness, influence, and opinion change in multi-agent systems that captures empirically observed deliberation patterns. We show that the FJ parameters are input-dependent, turning multi-agent deliberation into a mixture of experts. This perspective implies that multi-agent systems can outperform single agents and static ensembles when routing reflects agent competence. Since competence is latent in practice, we analyze how influence is established through observable proxies: agents' self-assessed confidence, their perceived confidence, and initial alignment with other agents' views.
