PART 02 / Beginner

Biological neuroscience

What does a neuron do, and what does a spike leave out?

Guided illustrated lesson

What you will learn

Draw a neuron, explain membrane currents, and distinguish synaptic mechanisms from abstract weights.

Before you begin: Foundations: matter, mathematics & computation

START WITH A QUESTION

Picture this.

A doorbell can ring five times politely or five times in a frantic burst. The count is the same; the timing is not. Neurons raise a similar question: what information do we lose when we replace a sequence of events with one number?

Spike train
The sequence of spike times for a neuron.
Depolarization
A shift toward a less negative membrane potential.
Decoder
A rule that turns recorded activity into an estimate or decision.

Read for the idea first. Work through the numbers next. Try answering the questions before opening the explanations.

Follow one signal through a cell

The familiar neuron drawing has branched dendrites, a cell body and an axon. Dendrites often receive synaptic input; the axon carries signals toward other cells. Treat this as an orientation map, not a universal blueprint—fly neurons can have quite different shapes. A membrane separates the inside from the outside. Channels provide selective ion pathways, while ion gradients supply part of the driving force. Electrical attraction and concentration differences both matter. Opening a channel changes a pathway; it does not simply pour in a substance called electricity.

From a small change to a spike

Depolarization means the membrane potential becomes less negative relative to its reference. Some voltage changes are small and graded. An action potential involves regenerative channel dynamics: a change can trigger processes that amplify it, followed by recovery. A timestamp model keeps when a spike happened while leaving out its waveform. Think of writing the times a doorbell rang instead of saving its sound recording. That is useful for some questions and inadequate for others. A refractory period describes reduced responsiveness after a spike.

A synapse is an interaction

At a chemical synapse, released transmitter interacts with receptors on another cell. The effect depends on the receptor, membrane state and timing. Excitatory effects tend to promote firing in the relevant context; inhibitory effects can oppose it or reduce responsiveness. In an artificial network we often compress all this into a weight. A negative number in a model is a convenient rule, not a complete account of inhibition. Always ask what biological detail the simplification keeps and what question it can answer.

Count, time and population

A spike train is a list of event times. Divide its count by the observation duration to obtain an average rate. Five spikes in 100 ms means 50 spikes per second on average; it does not promise one spike every 20 ms. A downstream circuit may respond differently to a burst than to evenly spaced events. Several neurons together also form a population pattern. Before choosing a representation, ask what behavior you want to explain: a rapid response, a sustained signal, or a combination of cells.

SEE THE RELATIONSHIPS

From structure to behavior.

Neuron: a labelled conceptual diagram
Figure 2.1 · Neuron. Original educational diagram; synthetic, not a measured result.
Spike train: a labelled conceptual diagram
Figure 2.2 · Spike train. Original educational diagram; synthetic, not a measured result.
The reasoning pathIon gradients then Membrane dynamics then Spike & synapse then Population activity1. Ion gradients2. Membrane dynamics3. Spike & synapse4. Population activity
A learning flow for this part. Follow the assumptions through to an observable result.

MAKE THE IDEA PRECISE

The mathematics.

r = N / T
Symbols and units
SymbolMeaning
rAverage firing rate, spikes per second (Hz)
NNumber of observed spikes
TObservation duration, seconds

Build the equation, step by step

  1. Choose the observation window before counting. A count without a duration cannot define a rate.
  2. Divide the spike count N by window length T in seconds. The resulting unit is spikes per second, conventionally reported in hertz.
  3. Changing the window changes the estimate. A longer window reduces time resolution; a short window can be dominated by whether a single spike fell just inside or outside its boundary.

WORK THROUGH THE NUMBERS

Worked example

Five spikes in 0.1 seconds give 50 Hz. This does not mean spikes are exactly 20 ms apart. Two trains can have identical counts and very different event timing.

A small Python experiment.

Python 3 · Standard library only · Synthetic educational example. Run this in your own notebook or a local file.

spikes_ms = [4, 6, 20, 70, 98]
window_ms = 100
rate_hz = len(spikes_ms) / (window_ms / 1000)
intervals = [b-a for a,b in zip(spikes_ms, spikes_ms[1:])]
print(f"rate = {rate_hz:.0f} Hz; intervals = {intervals} ms")

Read the code as a story

The list stores times in milliseconds. len counts events, and dividing the window by 1000 converts it to seconds. zip pairs each event with the next one; b-a gives the gaps [2, 14, 50, 28]. Those gaps reveal the burst that the rate alone hides. The first event's time is not itself an inter-spike interval unless a previous event is specified.

Expected output

rate = 50 Hz; intervals = [2, 14, 50, 28] ms

Change one quantity, predict the result, and then run the code again. Explain any difference before adding complexity.

Test your understanding.

Try it: Create a second five-spike train with evenly spaced events. Compare rates and inter-spike intervals.

Self-check: Can equal firing rates imply different messages?

Yes. Equal counts in equal windows give equal average rates but may have different spike timing, intervals or correlations with other cells. Whether those differences carry useful information depends on the downstream mechanism and task; timing is not automatically meaningful.

Where this model stops: A firing rate is a summary, not a complete neural code. Do not identify artificial activation values with measured biological signals without defining the mapping.

A research question: For a rapid reflex, when does a spike-count window become too slow?

Record evidence in your mastery worksheet →

PREDICT / TRY / EXPLAIN

Same rate, different rhythm

Use paper, a calculator or the Python example. Write your prediction before checking the result.

  1. Draw a 100 ms timeline and mark the five events from the example.
  2. Draw a second train at 10, 30, 50, 70 and 90 ms. Calculate its count, rate and four intervals.
  3. Use a toy detector that responds when two consecutive spikes are at most 5 ms apart. Apply it to both trains.
Compare with a worked solution

Both rates are 50 Hz. The evenly spaced train has four 20 ms intervals. Only the original train activates the toy detector because its first two spikes are 2 ms apart. This proves that a timing-sensitive rule can distinguish these examples. It does not prove that a particular biological circuit uses that rule.

CLOSE YOUR NOTES

Can you explain it back?

These questions test the reasoning, not just a remembered phrase. Say your answer out loud or write a sentence first.

01Three spikes arrive in 0.2 seconds. What is the rate?

Reveal the reasoning for question 1

3/0.2 = 15 Hz. Dividing by 200 without converting milliseconds would give the wrong unit.

02A pixel event and a neural spike both have timestamps. Are they the same thing?

Reveal the reasoning for question 2

No. Similar data formats can describe different mechanisms. Later we will connect event-camera reports to a neural model through an explicit encoding rule.

TAKE THE IDEA WITH YOU

Three things to remember.

  • A spike timestamp is a deliberate reduction of a biological event.
  • Rate depends on the chosen window; equal rates can hide different timing.
  • Artificial weights summarize interactions rather than reproducing all synaptic biology.

A cell becomes useful in a circuit. Next, we ask how connections combine signals and carry information through time.

Continue to Part 3 →

THE NEXT LAYER

Chapter depth map.

This introduction provides a first model. The following chapters are outlines for deeper future editions; their full treatments have not yet been published.

Cells and signals

2.1 A cell built for communication · Outline
  1. 2.1.1 Cells, membranes and intracellular compartments
  2. 2.1.2 Dendrites, soma, axon and axon hillock
  3. 2.1.3 Ion concentrations and selective channels
  4. 2.1.4 Resting potential and electrochemical gradients
  5. 2.1.5 Glia and the limits of a neuron-only account
2.2 From voltage to action potential · Outline
  1. 2.2.1 Depolarization, repolarization and hyperpolarization
  2. 2.2.2 Sodium and potassium conductance changes
  3. 2.2.3 Action potential initiation and propagation
  4. 2.2.4 Absolute and relative refractory periods
  5. 2.2.5 Graded potentials and species-specific physiology

Communication and adaptation

2.3 Synapses and plasticity · Outline
  1. 2.3.1 Chemical and electrical synapses
  2. 2.3.2 Calcium, transmitter release and receptors
  3. 2.3.3 Excitatory and inhibitory effects depend on reversal potentials
  4. 2.3.4 Neuromodulation, Hebbian learning and STDP
  5. 2.3.5 Homeostasis and plasticity across timescales
2.4 Neural codes and measurements · Outline
  1. 2.4.1 Spike frequency and rate coding
  2. 2.4.2 Spike timing and temporal coding
  3. 2.4.3 Population codes and decoding
  4. 2.4.4 Synchronization and correlated activity
  5. 2.4.5 Recording methods, measurement noise and interpretation

Read with the original sources.

These readings support the concepts and deeper investigation. The worked numbers and illustrations above are teaching examples.

Complete source register →

Where next?