HormoneHub: Modular Hormonal Modulation for Multi-Agent Cognitive Systems

This work introduces HormoneHub, a modular hormonal regulation framework designed to enhance dynamic behavior modulation in large language model (LLM)-based agents. Inspired by neurochemical systems in biological organisms, HormoneHub implements five core hormones — dopamine, serotonin, GABA, norepinephrine, and melatonin — each mapped to specific machine learning (ML) behavioral analogs. The system enables runtime modulation of generation temperature, memory retention, attention focus, cognitive inhibition, and latency. Receptors dynamically respond to hormone levels, translating biochemical signals into behaviorally meaningful parameter adjustments. HormoneHub serves as a foundation for embodied alignment, cognitive plasticity, and affective modulation in agent-based architectures.