MIRAGE: A Metric-Intensive Benchmark for Retrieval-Augmented Generation Evaluation
Paper • 2504.17137 • Published
Dataset Viewer
The dataset viewer is not available for this dataset.
Unexpected token '<', "<html>
<h"... is not valid JSON
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
MIRAGE dataset
MIRAGE is a benchmark dataset for evaluating Retrieval-Augmented Generation (RAG) systems, featuring 7,560 QA pairs and 37,800 context pools curated from diverse Wikipedia-based QA datasets (IfQA, NaturalQA, TriviaQA, DROP, PopQA). MIRAGE enables robust assessment of LLMs and retrievers under realistic, noisy, and oracle settings, and introduces novel metrics for analyzing context sensitivity, noise vulnerability, and retrieval effectiveness.
You can find our paper and github repository below:
- Downloads last month
- 82