Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.04790 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908753449713664 |
|---|---|
| author | Choi, Junhyuk Kwon, Jeongyoun Kim, Heeju Cho, Haeun Jung, Hayeong Min, Sehee Kim, Bugeun |
| author_facet | Choi, Junhyuk Kwon, Jeongyoun Kim, Heeju Cho, Haeun Jung, Hayeong Min, Sehee Kim, Bugeun |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_04790 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Belief in Authority: Impact of Authority in Multi-Agent Evaluation Framework Choi, Junhyuk Kwon, Jeongyoun Kim, Heeju Cho, Haeun Jung, Hayeong Min, Sehee Kim, Bugeun Computation and Language Artificial Intelligence 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. |
| title | Belief in Authority: Impact of Authority in Multi-Agent Evaluation Framework |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2601.04790 |