Large Language Model (LLM)

A large language model (LLM) is a neural network, usually a transformer, trained on very large text collections to predict the next token, which enables it to generate text, answer questions, summarise, translate and write code.

LLMs are pre-trained with self-supervised learning on large corpora of text and code, learning to predict the next token in a sequence. They are then adapted with instruction tuning and preference-based methods such as reinforcement learning from human feedback so that they follow instructions and respond helpfully. At inference time the model generates text one token at a time, sampling from predicted probabilities, with settings such as temperature controlling randomness.

Organisations use LLMs for chat assistants, document search and summarisation, report drafting, code generation, customer support, translation and natural-language interfaces to data and systems. Combined with retrieval-augmented generation and function calling, they can answer questions from company documents and act through tools, forming the core of AI agents.

LLMs can produce fluent but incorrect statements, known as hallucinations, have a limited context window and a knowledge cut-off, and can be manipulated through prompt injection. Outputs need grounding, evaluation and guardrails, and the handling of sensitive data must be considered when using hosted models. Model size, latency and cost drive the choice between hosted, self-hosted and smaller edge-deployable models.

Key points

Transformer models pre-trained to predict the next token on large text corpora
Instruction tuning and RLHF make them follow instructions
Prone to hallucination and limited by context window and knowledge cut-off
Combined with RAG and tools to build assistants and agents

Where AiVibe comes in

AiVibe delivers AI and machine learning services, chatbots and virtual assistants with RAG, MCP tools and voice, AI quality management including bias detection and model validation, and the AIMURUGA AI agent, and builds Intel-based edge AI devices using the Intel Distribution of OpenVINO toolkit.

Explore AiVibe’s work in AI & Machine Learning →

Large Language Model (LLM)

Key points

Where AiVibe comes in

Related terms

Terms that refer to Large Language Model (LLM)