What is Large Language Model (LLM)

Definition

A large language model (LLM) is a neural network trained on vast amounts of text to predict and generate language, enabling it to answer questions, summarise, translate, write, and reason over text based on patterns learned from its training data.

« Back to Glossary Index

Handles a wide range of language tasks without task-specific programming
Generates fluent, context-aware text from natural-language prompts
Adapts to new tasks through prompting rather than retraining
Serves as a flexible foundation for chatbots, assistants, and copilots

Real World Example

A support team integrates an LLM that reads a customer's question and the knowledge base, then drafts an accurate, on-brand reply for an agent to review, cutting average handling time substantially.

FAQs

How does an LLM generate text?

It predicts the most likely next token given the preceding context, repeating this to produce coherent passages of text.

What does 'large' refer to?

It refers to the model's scale, the billions of parameters and the massive text corpus it is trained on.

What are the limitations of LLMs?

They can produce plausible but incorrect answers, lack up-to-date or private knowledge unless supplied, and reflect biases in training data.