Shows how raw text is broken down into smaller units called "tokens" (words or sub-words) and converted into numerical IDs that the model can understand.