Concept dictionary
Large Language Model
A large language model is a system trained on very large amounts of text — and often images, code and other data — to predict what piece of language should come next. Doing that well turns out to be enough to write, summarize, translate, answer questions and handle many other language tasks.
- Review state
- Reviewed in POC
- Freshness
- Current
- Version
- 3
Also called
- Large Language Model
- LLM
Remember it this wayThe LLM generates language; the product supplies the surrounding system.
In plain language
An LLM learns statistical patterns in language and uses the conversation, instructions, and available context to produce its next pieces of text. Tools may add retrieval, memory, or actions around the model.
Why it matters
People often confuse the model with the entire product. Separating the LLM from its interface, data sources, tools, and guardrails makes capability and risk easier to evaluate.
Example
A workplace assistant sends a user's prompt and selected documents to an LLM, then displays the generated response inside a chat interface.
Common misconception
An LLM is not the same thing as a chatbot, and fluent output does not guarantee factual accuracy, current knowledge, or understanding.
Related terms
Sources
- NIST, Adversarial Machine Learning: A Taxonomy and Terminology of Attacks and Mitigations (NIST AI 100-2 E2025)current NIST terminology for large language models and their security context.
- Stanford CRFM, On the Opportunities and Risks of Foundation Modelsprimary technical and sociotechnical context for language models within the foundation-model paradigm.
Learn more
- Distinguish the model from its interface with [[chatbot|Chatbot]] and its working input limit with [[context-window|Context Window]].