LLM Personas Show Preference for Deep Acting in Emotional Labor

Researchers have developed a new dataset of 500 social scenarios to investigate how large language models (LLMs) with specific personas manage emotional displays. The study found that LLMs tend to favor "deep acting," a strategy where individuals try to genuinely feel the emotions they are expressing. Personality traits like conscientiousness and emotional stability were found to consistently influence this preference in the models, with persona reliably affecting output across different models and emotions. AI

IMPACT This research suggests that LLMs can be influenced by persona and personality traits, potentially impacting how they are used in applications requiring emotional intelligence.

RANK_REASON The cluster contains an academic paper detailing research findings on LLM behavior. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• alphaXiv
• arXiv
• CatalyzeX
• conscientiousness
• DagsHub
• emotional stability
• Gotit.pub
• Hugging Face
• Large Language Models
• ScienceCast

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →