Llms

  1. 1.What an LLM Actually Is (Not Another Prompting Course)
  2. 2.Subword Tokenization: Byte-Pair Encoding
  3. 3.Embeddings & Positional Encoding at Scale
  4. 4.Self-Attention, Formalized: Query, Key & Value
  5. 5.Multi-Head Attention & the Full Transformer Block
  6. 6.Decoder-Only vs. Encoder-Decoder: GPT vs. BERT vs. T5
  7. 7.Pretraining at Scale: Self-Supervised Learning
  8. 8.Scaling Laws
  9. 9.Fine-Tuning & RLHF: From GPT-3 to ChatGPT
  10. 10.Context Windows, Quadratic Attention & Honest Limitations
  11. 11.Capstone: Tracing a Prompt Through a Real LLM, End to End