Persona Agent
Persona agents are AI systems designed to interact and respond according to a defined personality or persona, enhancing their adaptability and relevance across various applications. Current research focuses on improving the accuracy and consistency of persona adherence, developing robust evaluation frameworks to measure performance, and exploring methods for generating diverse and persuasive outputs, often leveraging large language models (LLMs) and multi-agent frameworks. This field is significant because it advances the development of more engaging and personalized AI interactions in diverse sectors like education, healthcare, and entertainment, while also raising important questions about the reliability and security of these increasingly sophisticated agents.
Papers
AgentOhana: Design Unified Data and Training Pipeline for Effective Agent Learning
Jianguo Zhang, Tian Lan, Rithesh Murthy, Zhiwei Liu, Weiran Yao, Ming Zhu, Juntao Tan, Thai Hoang, Zuxin Liu, Liangwei Yang, Yihao Feng, Shirley Kokane, Tulika Awalgaonkar, Juan Carlos Niebles, Silvio Savarese, Shelby Heinecke, Huan Wang, Caiming Xiong
CloChat: Understanding How People Customize, Interact, and Experience Personas in Large Language Models
Juhye Ha, Hyeon Jeon, DaEun Han, Jinwook Seo, Changhoon Oh