___
[o_o]
/|___|\
|___|
mod-e is a neuromorphic mood simulation engine and visualizer. It demonstrates the real-time emotional state of a simulated brain using a smart robot face rendered in Raylib. The core of the engine is a spiking neural network that models emotional responses through the lens of the Lövheim Cube of Emotion.
By mapping the continuous release and decay of three primary neurotransmitters—Serotonin, Dopamine, and Noradrenaline—onto a 3D coordinate system (x, y, z), the engine dynamically transitions the robot's facial expressions across a spectrum of biologically grounded moods (e.g., Joy, Distress, Anger, Surprise).
The primary objective of this project is to provide a tangible, visual demonstration of a neuromorphic architecture. Instead of relying on traditional state machines for behavior, mod-e relies on internal stimuli, neurotransmitter modulation, and synaptic plasticity.
A key mechanism driving this system is Reward-Modulated Spike-Timing-Dependent Plasticity (R-STDP). Through R-STDP, the network doesn't just react to immediate stimuli; it reinforces pathways based on the dopamine-like reward signals generated by the simulated environment. This allows the simulated brain to adapt its emotional responses over time, demonstrating how biological learning mechanisms (R-STDP) can influence and shape complex emotional states defined by the Lövheim Cube.