Results 371 to 380 of about 2,337,590 (401)

Two Tales of Persona in LLMs: A Survey of Role-Playing and Personalization

Conference on Empirical Methods in Natural Language Processing
The concept of persona, originally adopted in dialogue literature, has re-surged as a promising framework for tailoring large language models (LLMs) to specific context (e.g., personalized search, LLM-as-a-judge).
Yu-Min Tseng, Yu-Chao Huang, Teng-Yun Hsiao, Yu-Ching Hsu, Jia-Yin Foo, Chao-Wei Huang, Yun-Nung Chen   +6 more
semanticscholar   +1 more source

Personas and Persons - An Empirical Study on Stereotyping of Personas [PDF]

Proceedings of Mensch und Computer 2019, 2019
Personas are considered a key method to induce empathy for users. However, they are also said to trigger stereotyping, e.g. by using photographs that convey information about gender, ethnicity, or age of the persona. To investigate this assumption, we conducted an eye-tracking study with 93 students from degree programs in computer science.
Pröbster, Monika, Hermann, Julia, Marsden, Nicola   +2 more
openaire   +2 more sources

PersonaGym: Evaluating Persona Agents and LLMs

arXiv.org
Persona agents, which are LLM agents conditioned to act according to an assigned persona, enable contextually rich and user aligned interactions across domains like education and healthcare. However, evaluating how faithfully these agents adhere to their
Vinay Samuel, Henry Peng Zou, Yue Zhou, Shreyas Chaudhari, A. Kalyan, Tanmay Rajpurohit, A. Deshpande, Karthik Narasimhan, Vishvak Murahari   +8 more
semanticscholar   +1 more source

PERSONA: A Reproducible Testbed for Pluralistic Alignment

International Conference on Computational Linguistics
The rapid advancement of language models (LMs) necessitates robust alignment with diverse user values. However, current preference optimization approaches often fail to capture the plurality of user opinions, instead reinforcing majority viewpoints and ...
Yuntao Bai, Andy Jones, Kamal Ndousse, Anna Askell, Dawn Chen, Stanislav Drain, Fort, Su Lin Blodgett, Solon Barocas, Hal Daumé, Louis Castricato, nathan lile, Suraj Anand, Souradip Chakraborty, Jiahao Qiu, Hui Yuan, Alec Kop-681, Furong Huang, Dinesh Manocha, Amrit Singh, Bedi Mengdi, Wang, Maxmin-rlhf, Jiangjie Chen, Xintao Wang, Rui Xu, Siyu Yuan, Wei Zhang, Jian Shi, Shuang Xie, Ruihan Li, Yang, Tinghui Zhu, Aili Chen, Nianqi Li, Caiyu Lida Chen, Siye Hu, Scott Wu, Ziquan Ren, Yanghua Fu, Ganqu Cui, Lifan Yuan, Ning Ding, Guanming Yao, Wei Zhu, Yuan Ni, Guotong Xie, Zhiyuan Liu, Amanda Schiefer, Anton Askell, Bakhtin, Carol, Zac Chen, Danny Hatfield-Dodds, Hernandez, Jan-Philipp Fränken, Sam Kwok, Peixuan Ye, Kanishk Gandhi, Dilip Arumugam, Jared Moore, Alex Tamkin, Tobias Gerstenberg, E. Zelikman, Rafael Rafailov, Kanishk Gandhi, Noah D Goodman. 2024, Hannah Rose Kirk, Alexander Whitefield, Paul Röttger, Andrew M. Bean, Katerina Margatina, Juan Ciro, Rafael Mosquera, Max Bartolo, Adina Williams, Yaswanth Chittepu, Ryan Park, Harshit S. Sikchi, Joey Hejna, Bradley Knox, Chelsea Finn, Shyam Sundhar Ramesh, Yifan Hu, Iason Chaimalas, Luca Rettenberger, Markus Reischl, Mark Schutera, Shibani Santurkar, Esin Durmus, Faisal Ladhak, Cinoo Lee, Percy Liang, Tatsunori Hashimoto, Omar Shaikh, Michelle Lam, Yijia Shao, Anand Siththaranjan, Cassidy Laidlaw, Taylor Sorensen, Jillian R. Fisher, Mitchell Gordon, Niloofar Mireshghallah, Michael Rytting, Andre Ye, Liwei Jiang, Ximing Lu, Nouha Dziri, Tim Althoff, Nisan Stiennon, Ouyang Long, Jeffrey Wu, Ryan Ziegler, Chelsea Lowe, Alec Voss, Radford, Dario Amodei, Christiano. 2020. Learn-860, Chenkai Sun, Revanth Ke Yang, Gangi Reddy, Hou Pong Chan, Chengxiang Zhai   +122 more
semanticscholar   +1 more source

Evaluating Large Language Model Biases in Persona-Steered Generation

Annual Meeting of the Association for Computational Linguistics
The task of persona-steered text generation requires large language models (LLMs) to generate text that reflects the distribution of views that an individual fitting a persona could have.
Andy Liu, Mona T. Diab, Daniel Fried
semanticscholar   +1 more source

Character is Destiny: Can Large Language Models Simulate Persona-Driven Decisions in Role-Playing?

arXiv.org
Can Large Language Models (LLMs) substitute humans in making important decisions? Recent research has unveiled the potential of LLMs to role-play assigned personas, mimicking their knowledge and linguistic habits.
Rui Xu, Xintao Wang, Jiangjie Chen, Siyu Yuan, Xinfeng Yuan, Jiaqing Liang, Zulong Chen, Xiaoqing Dong, Yanghua Xiao   +8 more
semanticscholar   +1 more source

A systematic narrative review of agent persona on learning outcomes and design variables to enhance personification

Journal of Research on Technology in Education, 2021
Early researchers suggested that virtual humans (animated pedagogical agents) required a persona if they were to be anthropomorphized by the user. Over the past two decades, research has focused on design features that target the perception of persona in
Robert O. Davis, Taejung Park, Joseph Vincent   +2 more
semanticscholar   +1 more source

Building Better AI Agents: A Provocation on the Utilisation of Persona in LLM-based Conversational Agents

International Conference on Conversational User Interfaces
The incorporation of Large Language Models (LLMs) such as the GPT series into diverse sectors including healthcare, education, and finance marks a significant evolution in the field of artificial intelligence (AI).
Guangzhi Sun, Xiao Zhan, Jose Such
semanticscholar   +1 more source

Evaluation of GenderMag Personas Based on Persona Attributes and Persona Gender

2016
We examined personas to see if (a) gendered persona attributes and (b) the persona presentation, i.e. the randomly assigned gender, would affect the perception of the persona. As stimulus material we used the GenderMag personas that were developed to inform the design process regarding empirically validated gender differences.
Nicola Marsden, Maren Haag
openaire   +2 more sources

PICLe: Eliciting Diverse Behaviors from Large Language Models with Persona In-Context Learning

International Conference on Machine Learning
Large Language Models (LLMs) are trained on massive text corpora, which are encoded with diverse personality traits. This triggers an interesting goal of eliciting a desired personality trait from the LLM, and probing its behavioral preferences ...
Hyeong Kyu Choi, Yixuan Li
semanticscholar   +1 more source