LLM · Transformers
LLM & Transformer whiteboards
A full public series from tokens to transformer stacks — teachable frames for machine learning and large language models.
- LLM & Transformers course map — teaching whiteboardCourse map for LLMs and Transformers: tokens, prediction, attention, and the full stack — a free public chalkline whiteboard.
- Next-token prediction — how LLMs generate textPredict the next token: the core of language models, explained on a free teaching whiteboard with worked examples.
- Word & token embeddings explained for LLMsWhat embeddings are and why language models need them — free public ML teaching whiteboard.
- RNNs for sequences — sequential NLP modelsRecurrent networks for language: how they read sequences and where they struggle — public teaching board.
- Attention as soft lookup — LLM building blockAttention as a soft dictionary lookup over values — concrete numbers on a free teaching whiteboard.
- Causal attention in language modelsWhy GPT-style models use causal (masked) attention so tokens only see the past — public whiteboard.
- Positional encodings in TransformersHow Transformers know token order without recurrence — positional encodings on a public teaching board.
- Train vs generate — how language models runTraining time vs generation time for language models — a clear public teaching whiteboard.
- LLM sampling: temperature, top-k, and top-pHow sampling turns a next-token distribution into text — temperature and nucleus sampling on a public whiteboard.
- Why attention beats CNNs and RNNs for languageThe motivation for attention: what CNNs and RNNs miss on long sequences — free LLM whiteboard.
- Scaled dot-product & multi-head attentionScaled dot-product attention and multi-head attention — the math behind Transformers on a free board.
- Transformer architecture stack explainedThe full Transformer stack: layers, residuals, and feed-forward blocks — free LLM whiteboard.
- From language model to AI assistantHow base LMs become assistants: instruction tuning and chat formatting — free public whiteboard.
- CNNs for sequences — NLP before TransformersUsing convolutional networks on text sequences: strengths and limits, taught on a free public whiteboard.
- LLM tokenization explained — from text to tokensHow language models turn text into tokens. Free public LLM teaching whiteboard with concrete string examples.
- Self-attention explained for TransformersSelf-attention: every token looks at every token. Free public Transformer teaching whiteboard.
More topics: Python · LLM / Transformers · Raspberry Pi · All public boards · Start free on chalkline