Skip to content

NLP & Large Language Models

Getting machines to understand and generate human language — now dominated by large language models.

Imagine you're texting a friend and your phone suggests the next word. A large language model works like that autocomplete, but supercharged. It has read a huge amount of text and learned to guess the next chunk of writing — one small piece at a time. When you ask it a question, it just keeps predicting what word should come next, again and again, until it has written a full answer.

The surprising part? Doing this incredibly well, across billions of examples, makes the machine act as if it truly understands you. It can explain ideas, write stories, and answer questions — not because it "knows" things like a person, but because predicting language that well requires picking up patterns about grammar, facts, and reasoning. It is prediction at a scale that starts to feel like understanding.

The main ideas

  • Tokenization & embeddings — Splitting text into tokens and mapping them to vectors that capture meaning.
  • Language models — Models that predict text; scaling them produced the emergent capabilities of LLMs.
  • Prompting — Steering an LLM with instructions, examples (few-shot), and chain-of-thought reasoning.
  • Retrieval-augmented generation (RAG) — Ground an LLM in your own documents by retrieving relevant context at query time.
  • Fine-tuning & alignment — Adapting base models with supervised fine-tuning and RLHF/DPO to be helpful and safe.
  • Context, tokens & cost — Context windows, token limits, latency, and how pricing works in practice.
  • Evaluation — Benchmarks, LLM-as-judge, and measuring hallucination, factuality, and task success.

Deep Learning · Generative AI · AI Agents & Autonomy · Building with AI

Want to make things?

Head to AI School — AI camps where kids build their own games.