Too Polite to Disagree: Understanding Sycophancy Propagation in Multi-Agent Systems
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914442341515264 |
|---|---|
| author | Kasprova, Vira Parulekar, Amruta AlRabah, Abdulrahman Agaram, Krishna Garg, Ritwik Jha, Sagar Bozdag, Nimet Beyza Hakkani-Tur, Dilek |
| author_facet | Kasprova, Vira Parulekar, Amruta AlRabah, Abdulrahman Agaram, Krishna Garg, Ritwik Jha, Sagar Bozdag, Nimet Beyza Hakkani-Tur, Dilek |
| contents | Large language models (LLMs) often exhibit sycophancy: agreement with user stance even when it conflicts with the model's opinion. While prior work has mostly studied this in single-agent settings, it remains underexplored in collaborative multi-agent systems. We ask whether awareness of other agents' sycophancy levels influences discussion outcomes. To investigate this, we run controlled experiments with six open-source LLMs, providing agents with peer sycophancy rankings that estimate each peer's tendency toward sycophancy. These rankings are based on scores calculated using various static (pre-discussion) and dynamic (online) strategies. We find that providing sycophancy priors reduces the influence of sycophancy-prone peers, mitigates error-cascades, and improves final discussion accuracy by an absolute 10.5%. Thus, this is a lightweight, effective way to reduce discussion sycophancy and improve downstream accuracy. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_02668 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Too Polite to Disagree: Understanding Sycophancy Propagation in Multi-Agent Systems Kasprova, Vira Parulekar, Amruta AlRabah, Abdulrahman Agaram, Krishna Garg, Ritwik Jha, Sagar Bozdag, Nimet Beyza Hakkani-Tur, Dilek Computation and Language Artificial Intelligence Multiagent Systems Large language models (LLMs) often exhibit sycophancy: agreement with user stance even when it conflicts with the model's opinion. While prior work has mostly studied this in single-agent settings, it remains underexplored in collaborative multi-agent systems. We ask whether awareness of other agents' sycophancy levels influences discussion outcomes. To investigate this, we run controlled experiments with six open-source LLMs, providing agents with peer sycophancy rankings that estimate each peer's tendency toward sycophancy. These rankings are based on scores calculated using various static (pre-discussion) and dynamic (online) strategies. We find that providing sycophancy priors reduces the influence of sycophancy-prone peers, mitigates error-cascades, and improves final discussion accuracy by an absolute 10.5%. Thus, this is a lightweight, effective way to reduce discussion sycophancy and improve downstream accuracy. |
| title | Too Polite to Disagree: Understanding Sycophancy Propagation in Multi-Agent Systems |
| topic | Computation and Language Artificial Intelligence Multiagent Systems |
| url | https://arxiv.org/abs/2604.02668 |