Differential equations are the common language of neuroscience. The same kind of equation describes a tiny ion channel, a single neuron, a whole cortical population, and the choice you make at a coffee shop. Learn the language once, read all of it.

Below are the famous ones. Don't memorize the names — just notice they line up by how zoomed-in they are: one cell → millions of cells → the whole person. Here's what each actually measures, in plain words.

🔬 Single-neuron models — zoomed all the way in to ONE cell.

They track the voltage inside a single neuron (the tiny electrical charge across its membrane) as it rises and falls, and the instant it fires a spike.

• Hodgkin–Huxley — the most detailed: it follows the actual microscopic gates (sodium and potassium 'doors' that pop open and shut) that physically build a spike. Most realistic, heaviest to compute.
• FitzHugh–Nagumo / Izhikevich — simpler cousins that drop most of the gate detail but still reproduce the spike's shape. Less realistic, fast to run.
• Leaky Integrate-and-Fire (LIF) — the simplest: it ignores the gates and just lets the voltage fill up like a bucket and reset every time the neuron fires. Easiest to learn, so it's the one we'll build in detail.

🧠 Population & whole-brain models — zoomed out to MILLIONS of neurons at once.

Tracking every neuron is hopeless, so these follow an average instead.

• Rate models / Wilson–Cowan / neural mass models — measure how busy a brain region is: what fraction of its neurons are active right now. Wilson–Cowan tracks two tugging crowds — the excitatory (gas pedal) and inhibitory (brake) groups — pushing and pulling on each other.
• Balloon model — measures blood flow and oxygen, not spikes. Why? An fMRI scanner can't see spikes; it only sees blood rushing to busy areas. This model turns 'neurons got active' into 'blood flowed in' — the colored blobs in brain-scan images.

🙋 Behavior & cognition models — zoomed all the way out to the whole PERSON.

No neurons here, just what someone does.

• Drift-Diffusion Model — pictures a decision as evidence piling up like sand on a scale until it tips past a finish line. It predicts both which choice you make and how long it takes (reaction time).
• Leaky Integrator of forgetting — pictures a memory as water in a leaky cup: it slowly drains away unless you keep topping it up. It predicts how fast you forget — like a phone number slipping away seconds after you hear it.

🤯 The fact worth remembering: that leaky cup of forgetting and the leaky bucket neuron (LIF) run on almost the same equation. One bit of math describes both a neuron's voltage leaking back down and a memory fading away. Nature reuses good math — which is exactly why learning this one tool unlocks so much.

🦑 Fun history — how the spike was actually cracked.

In the late 1940s, Alan Hodgkin and Andrew Huxley needed a nerve fiber big enough to experiment on. They found a monster in the squid: its giant axon — the nerve that triggers the squid's panic jet-escape — is about 1 mm thick, roughly 100× fatter than a human nerve fiber. Big enough to thread a thin wire right inside it.

Using a clever trick called a voltage clamp (hold the membrane voltage fixed, then measure the electric current that flows), they watched charged ions cross the membrane and worked out what a spike really is: first sodium ions rush IN and the voltage shoots up, then potassium ions rush OUT and it drops back down — through tiny gates that open and close in sequence. That in-then-out surge, in about a millisecond, is the spike.

The stunning part: they captured all of it in a set of differential equations that reproduced a spike's exact shape and speed — years before anyone could see those gates under a microscope. The math predicted the biology. It won them the 1963 Nobel Prize, and it's why differential equations sit at the foundation of neuroscience.

The one-sentence takeaway: a differential equation is a rule that tells you how a system changes over time. Instead of stating the future directly, it states the rate of change — and from that, the whole future unfolds.