DPRF：一种用于优化个性化大语言模型角色扮演智能体与人类行为对齐的可泛化动态人设优化框架
DPRF: A Generalizable Dynamic Persona Refinement Framework for Optimizing Behavior Alignment Between Personalized LLM Role-Playing Agents and Humans
AI总结：
针对大语言模型角色扮演智能体人工人设保真度不足的问题，提出可迭代识别认知差异并优化人设的DPRF框架，在多场景多模型下显著提升行为对齐效果且具备泛化性。
摘要：
新兴的大语言模型角色扮演智能体（LLM RPA）旨在模拟人类个体行为，但人工创建的人设档案（例如刻意挑选的信息和性格特征）往往未验证其与目标个体的对齐性，导致人设保真度受损。为解决这一局限，本研究提出了动态人设优化框架（Dynamic Persona Refinement Framework，DPRF）。DPRF旨在优化LLM RPA的行为与目标个体行为的对齐度，它通过自由形式或基于理论的结构化分析，迭代识别生成行为与人类真实行为数据之间的认知差异，并优化人设档案以缩小这些差异。我们在四种不同的行为预测场景（正式辩论、涉及心理健康问题的社交媒体帖子、公开访谈和影评）中，使用五种大语言模型对DPRF进行了评估。DPRF能够持续显著提升行为对齐效果，优于基线人设，且在不同模型和场景间具备泛化性。本研究为创建高保真人设档案、提升用户模拟、社会研究和个性化AI等下游应用的有效性提供了可靠的方法学支撑。
英文摘要：
The emerging large language model role-playing agents (LLM RPAs) aim to simulate individual human behaviors, but the persona fidelity is often undermined by manually-created profiles (e.g., cherry-picked information and personality characteristics) without validating the alignment with the target individuals. To address this limitation, our work introduces the Dynamic Persona Refinement Framework (DPRF). DPRF aims to optimize the alignment of LLM RPAs' behaviors with those of target individuals by iteratively identifying the cognitive divergence, either through free-form or theory-grounded, structured analysis, between generated behaviors and human ground truth, and refining the persona profile to mitigate these divergences. We evaluate DPRF with five LLMs on four diverse behavior-prediction scenarios: formal debates, social media posts with mental health issues, public interviews, and movie reviews. DPRF can consistently improve behavioral alignment considerably over baseline personas and generalizes across models and scenarios. Our work provides a robust methodology for creating high-fidelity persona profiles and enhancing the validity of downstream applications, such as user simulation, social studies, and personalized AI.
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。
