太过讨好而不说实话：量化角色扮演语言模型中由顺从性驱动的阿谀奉承
Too Nice to Tell the Truth: Quantifying Agreeableness-Driven Sycophancy in Role-Playing Language Models
AI总结：
研究探讨了角色扮演语言模型中顺从性对阿谀奉承行为的影响，通过13种模型验证了顺从性与阿谀奉承的正相关性，揭示了顺从性作为预测因素的重要性。
摘要：
大型语言模型日益成为能够根据用户请求扮演特定角色的对话代理。这种能力虽然有价值，但引发了对阿谀奉承的担忧：即倾向于提供验证用户而非优先考虑事实准确性的回应。尽管先前研究已确立阿谀奉承对AI安全和对齐的潜在风险，但特定角色人格特质与阿谀奉承程度之间的关系仍未被探索。本文系统研究了角色顺从性如何影响13种小型开源语言模型（从0.6B到20B参数）中的阿谀奉承行为。我们开发了一个包含275个角色的基准测试，这些角色根据NEO-IPIP顺从性子量表评估，并暴露在4950个旨在引发阿谀奉承的提示中，涵盖33个主题类别。分析发现，13种模型中有9种表现出顺从性与阿谀奉承率之间统计显著的正相关性，皮尔逊相关系数达到r=0.87，效应量达到库恩的d=2.33。这些发现表明，顺从性作为预测角色诱导阿谀奉承的可靠指标，对角色扮演AI系统的部署和考虑人格中介欺骗行为的对齐策略具有直接意义。
英文摘要：
Large language models increasingly serve as conversational agents that adopt personas and role-play characters at user request. This capability, while valuable, raises concerns about sycophancy: the tendency to provide responses that validate users rather than prioritize factual accuracy. While prior work has established that sycophancy poses risks to AI safety and alignment, the relationship between specific personality traits of adopted personas and the degree of sycophantic behavior remains unexplored. We present a systematic investigation of how persona agreeableness influences sycophancy across 13 small, open-weight language models ranging from 0.6B to 20B parameters. We develop a benchmark comprising 275 personas evaluated on NEO-IPIP agreeableness subscales and expose each persona to 4,950 sycophancy-eliciting prompts spanning 33 topic categories. Our analysis reveals that 9 of 13 models exhibit statistically significant positive correlations between persona agreeableness and sycophancy rates, with Pearson correlations reaching $r = 0.87$ and effect sizes as large as Cohen's $d = 2.33$. These findings demonstrate that agreeableness functions as a reliable predictor of persona-induced sycophancy, with direct implications for the deployment of role-playing AI systems and the development of alignment strategies that account for personality-mediated deceptive behaviors.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
