Pillar 1

Transformer & LLM Internals

explain how an LLM works end-to-end without deep math. 3Blue1Brown chapters are the visual gold standard - watch in order

Big-picture: what an LLM is
LLMs explained briefly - 3B1B (playlist) ↗video8m
Sign in to track this item
Full forward pass, embeddings, softmax
Ch.5 Transformers, the tech behind LLMs - 3B1B ↗video27m
Sign in to track this item
Attention: Q/K/V, masking, multi-head KEY
Ch.6 Attention, step-by-step - 3B1B ↗video26m
Sign in to track this item
How facts live in MLP layers (often skipped)
Ch.7 How might LLMs store facts - 3B1B ↗video23m
Sign in to track this item
Tokenization/BPE + positional encoding
Let's build GPT from scratch - Karpathy ↗video1h56m
Sign in to track this item