Chapter 3.1 - Tokenization & Input Prompt
Tokenization is the very first step. Tokenization serves as the foundational mechanism to transform human-readable text into small, structured pieces (like syllables or words) that a computer program can understand.
Chapter 3.2 - Token Embedding Lookup (We Matrix)
Token Embeddings are the magic trick that turns completely random ID numbers into rich, meaningful concepts that the AI can understand.
Chapter 3.3 - Positional Embeddings & Combined Input Matrix
Positional embeddings are how we teach the AI the order of words in a sentence, giving it a sense of time and sequence.