Emoti – A Large Language Model Supporting Emotion Regulation Based On Biodata

Engineering notes

Engineering notes will be added by the aipentium editorial team.

Chinese explanation / 中文解读

中文解读待补充：本站会优先为大语言模型、生成式AI、ChatGPT相关技术、计算机视觉、深度学习等高价值论文补充中文说明。

Original abstract

Adequately regulating one's emotions is important for the physical and mental well-being but requires practice.However, support remains in high demand.Here, chatbots can offer immediate assistance by mimicking a human interaction, as the high level of responsiveness of transformer-based large language models (LLMs) or generative pre-trained transformer (GPT) chatbots can create a sense of social presence.Nevertheless, most studies focus exclusively on bodily responses, neglecting strategy training.Hence, solutions are frequently rule-based and lack a foundation in adaptive artificial intelligence (AI).In our TREO paper, we introduce EMOTI, a multi-agent system based on GPT that helps regulate emotions at different levels.This includes learning about emotions and interacting with a so-called situative agent that evolves by using biodata from a wearable.Once our research is complete, we will contribute to theory and practice by delivering insights on LLMs, emotion regulation, and the integration of wearables and LLM data.

Links and sources

• PDF from original source

Comments