Belief in Authority: Impact of Authority in Multi-Agent Evaluation Framework
This addresses the underexplored impact of authority bias in multi-agent systems, providing insights for designing frameworks with asymmetric interactions, but it is incremental as it builds on existing theories and methods.
The study systematically analyzed role-based authority bias in multi-agent systems using ChatEval, finding that Expert and Referent power roles had stronger influence than Legitimate roles, with bias emerging from authoritative roles maintaining positions while general agents showed flexibility.
Multi-agent systems utilizing large language models often assign authoritative roles to improve performance, yet the impact of authority bias on agent interactions remains underexplored. We present the first systematic analysis of role-based authority bias in free-form multi-agent evaluation using ChatEval. Applying French and Raven's power-based theory, we classify authoritative roles into legitimate, referent, and expert types and analyze their influence across 12-turn conversations. Experiments with GPT-4o and DeepSeek R1 reveal that Expert and Referent power roles exert stronger influence than Legitimate power roles. Crucially, authority bias emerges not through active conformity by general agents, but through authoritative roles consistently maintaining their positions while general agents demonstrate flexibility. Furthermore, authority influence requires clear position statements, as neutral responses fail to generate bias. These findings provide key insights for designing multi-agent frameworks with asymmetric interaction patterns.