The artificial neuron is the elementary unit of a neural network. It takes inputs, computes a weighted sum, adds a bias, and passes it through an activation function.
Modeled after biological neurons (Rosenblatt's Perceptron).
The fundamental atom of all deep learning.