Scalar Accumulation
Key concept: A neuron sums weighted inputs and a bias: a = w₁x₁ + w₂x₂ + b.
Adjust w₁, w₂, and b for the fixed input x = (2, −1). Step through adds each term; red and blue bars show positive and negative contributions. The right panel shows the same sum geometrically.
What to look for: changing a weight moves both its bar and the vector w. Both views represent the same output.
Open the demo in a new tab for more room.