AI engineering glossary · Evals guardrails AI security
What is Llama Guard?
Llama Guard: An open safety classifier model from Meta, based on the Llama family, that reads a conversation and labels the user input or the model reply as safe or unsafe against a configurable list of harm categories. It is used as an input or output guardrail.
Llama Guard: an example
An application sends each user prompt and model answer to the classifier and blocks any labelled unsafe.
Related terms
More Evals guardrails AI security terms
- LLM-as-a-judge
- lm-evaluation-harness
- Many-shot jailbreaking
- Membership inference attack
- Misinformation risk
- MITRE ATLAS
- MMLU
- Model extraction
Where you learn it
Llama Guard is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.