Characterizing Conversational Tones in Humans and LLMs

• Objective: Explore similarities and differences in conversational tones between humans and LLMs.
• Method: Iterative elicitation of conversational tones and sentences from human participants and LLMs.
• Results: Develop a dataset for creating a geometric representation of tone relations.
• Significance: Enhances understanding of human-computer interactions with cognitive science insights.

This study highlights the importance of characterizing conversational tones and the potential for improving communication interfaces through machine learning approaches.