Generative Intelligence Systems in the Flow of Group Emotions

TL;DR

This paper presents a model for artificial agents to detect, infer, and generate emotional responses to influence group mood dynamics in real-time, advancing affective computing from individual to group-level applications.

Contribution

It introduces a novel system for orchestrating emotion contagion among artificial agents and humans, enabling coordinated group emotion modulation in digital interactions.

Findings

Effective detection of emotional signals in group settings

Successful generation of targeted emotional responses

Influence on group mood dynamics demonstrated

Abstract

Emotional cues frequently arise and shape group dynamics in interactive settings where multiple humans and artificial agents communicate through shared digital channels. While artificial agents lack intrinsic emotional states, they can simulate affective behavior using synthetic modalities such as text or speech. This work introduces a model for orchestrating emotion contagion, enabling agents to detect emotional signals, infer group mood patterns, and generate targeted emotional responses. The system captures human emotional exchanges and uses this insight to produce adaptive, generative responses that influence group affect in real time. The model supports applications in collaborative, educational, and social environments by shifting affective computing from individual-level reactions to coordinated, group-level emotion modulation. We present the system architecture and provide experimental results that illustrate its effectiveness in sensing and steering group mood dynamics.