# RAG Evaluation

> RAG evaluation measures a retrieval-augmented system's retrieval and generation separately, typically checking whether the retrieved context is relevant, whether the answer is faithful to that context, and whether the answer addresses the question.

- Identifier: PTL-0181
- Category: Failure Modes & Evaluation
- Canonical URL: https://protologue.com/t/rag-evaluation/
- Also known as: Ragas, RAG metrics, faithfulness, context relevance
- Introduced: 2023

## Description

Separating the stages shows whether a bad answer came from retrieval or from generation. Frameworks such as Ragas compute these metrics automatically with an LLM, without reference answers, which makes it practical to compare chunking, retrieval, and prompt choices.

## Broader terms

- [Evals](https://protologue.com/t/evals/)

## Related terms

- [Retrieval-Augmented Generation](https://protologue.com/t/retrieval-augmented-generation/)
- [LLM-as-a-Judge](https://protologue.com/t/llm-as-a-judge/)
- [Hallucination](https://protologue.com/t/hallucination/)
- [Chunking](https://protologue.com/t/chunking/)

## Sources

- Es et al. (2023). Ragas: Automated Evaluation of Retrieval Augmented Generation. https://arxiv.org/abs/2309.15217

## Cite this entry

Protologue. (2026). RAG Evaluation. In Protologue: A Taxonomy of Prompting and LLM Techniques (v1.4.0, PTL-0181). https://protologue.com/t/rag-evaluation/

License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)