Layer 28

Nervous Systems, Sensing & Neural Dynamics

Membrane integration, population coding, excitation–inhibition dynamics and plasticity show how living matter can process fast signals.

interactivemodel assumptions visibleestablished science separated from analogy

Story step: On some worlds, living matter builds fast electrical networks. Nervous systems transform gradients and signals into sensing, action and learning.

1 · Leaky integrate-and-fire neuron

A membrane behaves partly like a resistor–capacitor circuit. If voltage crosses a threshold in this reduced model, a spike is registered and the voltage resets.

τ dV/dt = −(V−E_L) + R I

input current I

membrane time τ

threshold

—spikes in window

—model firing rate

reduced modelnot a Hodgkin–Huxley action potential

2 · Population coding

A stimulus can be represented by the activity of many broadly tuned neurons rather than one dedicated detector.

stimulus angle θ

tuning sharpness κ

—decoded population vector

—angular error

—peak response

3 · Excitation–inhibition network dynamics

A Wilson–Cowan population model shows how coupled excitatory and inhibitory activity can settle, oscillate or saturate.

τE E˙ = −E + S(wEE E − wEI I + P) τI I˙ = −I + S(wIE E − wII I + Q)

external drive P

E→E coupling

I→E inhibition

E→I coupling

excitatory Einhibitory I

4 · Hebbian correlation learning

A minimal Hebbian rule strengthens a connection when presynaptic and postsynaptic activities are correlated.

Δw = η x y

input activity x

output activity y

learning rate η

—current weight

—last Δw

local rulenot a complete theory of biological learning

Scientific boundary: Leaky integrate-and-fire, tuning curves, Wilson–Cowan populations and Hebbian updates are reduced models. They do not reproduce all cellular physiology, learning mechanisms, cognition or consciousness.