What is an LLM?
A type of AI model trained on vast amounts of text to understand and generate human language.
Stands for: Large Language Model
Pronunciation: /ˌel el ˈem/
LLM explained in plain English
A Large Language Model (LLM) is an artificial intelligence system trained on enormous collections of text — books, websites, articles, and more. By studying patterns in this text, the model learns to predict what word or phrase should come next. This ability lets it answer questions, write essays, summarise documents, translate languages, and hold conversations.\n\nLLMs do not truly "understand" language the way humans do. Instead, they are extraordinarily sophisticated pattern-matching engines that produce text that often reads as if it were written by a knowledgeable person.
Analogy
Think of an LLM like a librarian who has read every book in the world but has never left the library. They can discuss almost any topic convincingly because they have seen so many examples, but their knowledge comes entirely from text — not from direct experience of the world.
Example
How is LLM used?
LLMs power chatbots like ChatGPT, coding assistants, search engines, content generation tools, and document analysis systems. Developers integrate them via APIs to add natural language capabilities to applications.
Common misconceptions about LLM
LLMs do not have consciousness or genuine understanding. They can produce confident-sounding but incorrect information ("hallucinations"). They also reflect biases present in their training data.
History
The transformer architecture (2017) enabled modern LLMs. GPT-2 (2019) demonstrated surprising text generation. GPT-3 (2020) showed few-shot learning at scale. ChatGPT (2022) brought LLMs to mainstream awareness.
Related terms
People also read
- Transformer
The neural network architecture that revolutionised AI by enabling models to process entire sequences at once.
- Token
The basic unit of text that AI language models process, which may be a word, part of a word, or punctuation.
- Embedding
A numerical representation of text, images, or other data that captures semantic meaning.
- Context Window
The maximum amount of text a language model can consider at one time, measured in tokens.
- Fine-tuning
The process of further training a pre-trained AI model on specialised data to improve performance on specific tasks.
- Prompt
The input text or instruction given to an AI model to guide its response.
- GPT
A family of large language models developed by OpenAI that generate human-like text.
- RAG
A technique that combines AI language models with external knowledge retrieval for more accurate answers.
- AI slop
Output from a generative AI system that favors quantity over quality.
- BERT
A model architecture for text representation.