多代理角色设置下,大模型出现的社会行为

Journal: Advances in Computer and Autonomous Intelligence Research DOI: 10.12238/acair.v2i2.7347

霍虹颖

北京语言大学

Abstract

本研究旨在探索多代理角色设置对大语言模型性能的影响,并深入分析在此设置下模型所表现出的社会行为。通过消融实验的设置、观察影响因素的改变发现,大语言模型在多代理交流和角色设置的场景下会产生丰富的社会行为,这样的设置不仅可以提升大语言模型的性能,还使其行为更加接近人类。本研究不仅为理解大语言模型在自然语言处理中的应用提供了新视角,也为后续研究提供了重要参考。

Keywords

大语言模型;多代理;角色扮演;社会行为

References

[1] Wei, J., Wang, X., Schuurmans, D., Bosma, M., **a, F., Chi,E.,...& Zhou,D.(2022).Chain-of-thought prompting elicits reaso ning in large language models.Advances in neural information processing systems,35,24824-24837.
[2] Chan,C.M.,Chen,W.,Su, Y., Yu, J., Xue, W., Zhang, S., ... & Liu, Z.(2023).Chateval: Towards better llm-based evaluators throu gh multi-agent debate.arxiv preprint arxiv:2308.07201.
[3] Tao,M.,Liang,X.,Shi,T.,Yu,L.,&**e,Y.(2023). RoleCraft-GLM: Advancing Personalized Role-Playing in Large Language Mod els.arxiv preprint arxiv:2401.09432.
[4] Chen, W., Su, Y., Zuo, J., Yang, C.,Yuan,C.,Chan,C.M.,...& Zhou,J.(2023, October). Agentverse: Facilitating multi-agent collaboration and exploring emergent behaviors.In The Twel fth International Conference on Learning Representations.
[5] Cialdini,R.B., & Goldstein, N. J. (2004). Social influence: Compliance and conformity.Annu.Rev.Psychol.,55,591-621.
[6] Zhou,Z.,Lin,Y.,**, D.,& Li,Y. (2024). Large Language Model for Participatory Urban Planning.arxiv preprint arxiv:2402. 17161.

Copyright © 2024 霍虹颖

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License