arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.03545cs.AI
收藏

人格生成器:为任意上下文生成多样化的合成人格

Persona Generators: Generating Diverse Synthetic Personas for Arbitrary Contexts

Google DeepMind(谷歌DeepMind)

Davide Paglieri, Logan Cross, William A. Cunningham, Joel Z. Leibo, Alexander Sasha Vezhnevets

AI总结:

提出Persona Generators,通过迭代进化优化生成覆盖广泛意见和偏好的多样化合成人格,在六个多样性指标上显著优于现有基线。

摘要:

评估与人类交互的AI系统需要理解它们在不同用户群体中的行为,但收集代表性人类数据通常成本高昂或不可行,特别是对于新技术或假设的未来场景。最近在生成式基于智能体建模方面的工作表明,大型语言模型可以高保真地模拟类似人类的合成人格,准确再现特定个体的信念和行为。然而,大多数方法需要关于目标群体的详细数据,并且通常优先考虑密度匹配(复制最可能的内容)而非支持覆盖(覆盖可能的内容),导致长尾行为未被充分探索。我们引入了Persona Generators,即能够为任意上下文生成多样化合成群体的函数。我们应用基于AlphaEvolve的迭代改进循环,使用大型语言模型作为变异算子,在数百次迭代中优化我们的Persona Generator代码。优化过程产生了轻量级的Persona Generators,能够自动将小规模描述扩展为多样化的合成人格群体,这些群体在相关多样性轴上最大化意见和偏好的覆盖。我们证明,进化后的生成器在保留上下文上的六个多样性指标上显著优于现有基线,产生了覆盖标准LLM输出中难以实现的罕见特征组合的群体。

英文摘要:

Evaluating AI systems that interact with humans requires understanding their behavior across diverse user populations, but collecting representative human data is often expensive or infeasible, particularly for novel technologies or hypothetical future scenarios. Recent work in Generative Agent-Based Modeling has shown that large language models can simulate human-like synthetic personas with high fidelity, accurately reproducing the beliefs and behaviors of specific individuals. However, most approaches require detailed data about target populations and often prioritize density matching (replicating what is most probable) rather than support coverage (spanning what is possible), leaving long-tail behaviors underexplored. We introduce Persona Generators, functions that can produce diverse synthetic populations tailored to arbitrary contexts. We apply an iterative improvement loop based on AlphaEvolve, using large language models as mutation operators to refine our Persona Generator code over hundreds of iterations. The optimization process produces lightweight Persona Generators that can automatically expand small descriptions into populations of diverse synthetic personas that maximize coverage of opinions and preferences along relevant diversity axes. We demonstrate that evolved generators substantially outperform existing baselines across six diversity metrics on held-out contexts, producing populations that span rare trait combinations difficult to achieve in standard LLM outputs.

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。

↑