Connectome

🎓 Lesson 1: Your First Neuron

Understanding the building block of neural networks

Your Challenge

Make the neuron's output match the target (1.0) by adjusting the inputs and weights!

How to Learn

1

Watch the Flow

See how Input A (1.0) and Input B (0.0) flow into the neuron. The w = 0.50 on the edges are the weights that control how much each input matters.

2

Examine the Neuron

Look at the purple neuron in the center. It shows you every calculation: multiply inputs by weights, add them together (weighted sum), add bias, and apply sigmoid activation.

3

Check the Loss

The Loss Calculator (far right) shows how far we are from the target. Lower loss = better! Try to get it close to 0.00.

4

Experiment!

Move input sliders, click edge weights to edit them, adjust bias in the neuron, and try different activation functions (sigmoid, tanh, relu).

Key Insights

•Weights control how much each input matters
•Bias shifts the neuron's tendency to activate
•Activation function (sigmoid) squashes output between 0 and 1
•To hit the target, you need the right combination of all three!

You've Learned This When

Get neuron output close to 1.0 (loss < 0.05)

Understand what each number in the neuron calculation means

Can predict what happens when you change a weight

Ready for More?

Once you understand one neuron, let's tackle a real problem that stumped AI researchers for years...

INPUT

1.00

-11.00+1

INPUT

0.00

-10.00+1

NEURON

Our Neuron

Output

0.622

Step 1: Weighted Inputs

\sum (w_i \times x_i)

Σ = 0.5000

Step 2: Add Bias

Bias (b):0.00

z = 0.500 + 0.00

z = 0.5000

Step 3: Activation Function

Squashes values between 0 and 1. Great for probabilities!

\sigma(z) = \frac{1}{1 + e^{-z}}

σ(0.500)

0.6225

FINAL OUTPUT

0.6225

INPUT

1.00

-11.00+1

LOSS CALCULATOR

PREDICTED

0.500

ACTUAL

1.000

Error: -0.500

CALCULATION:

Error = Predicted - Actual

Error = 0.500 - 1.000 = -0.500

Squared Error = (-0.500)²

Squared Error = 0.250000

Loss (MSE) = 0.250000

LOSS

0.250000

ACTIVATION FUNCTION

σ(x) = 1 / (1 + e^(-x))

Input Value:

-6x = 0.00+6

OUTPUT

0.5000

First Neuron