🧠 3D Leaky Integrate-and-Fire Spiking Lattice: Exact f-I Curve, Event-Driven versus Clocked Updates and Operation Count
A Blender-built lattice of 343 pearl beads (7x7x7 leaky integrate-and-fire neurons, 20 percent inhibitory) lights up with spikes: drive it with Poisson input, kick the centre, and compare event-driven with clocked updating by exact operation counts, with the closed-form f-I curve, Euler order and Campbell shot-noise checks.
⚙ Under the hood
A Blender-built lattice of 343 pearl beads (7x7x7 leaky integrate-and-fire neurons, 20 percent inhibitory) lights up with spikes: drive it with Poisson input, kick the centre, and compare event-driven with clocked updating by exact operation counts, with the closed-form f-I curve, Euler order and Campbell shot-noise checks.
Spiking Neural NetworkLeaky Integrate-and-FireNeuromorphic ComputingEvent-Driven Simulationf-I CurvePoisson InputComputational NeuroscienceEnergy Efficiency
3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install