BEIR

Dataset Card for BEIR Benchmark Dataset Summary BEIR is a heterogeneous benchmark that has been built from 18 diverse datasets representing 9 information retrieval tasks: Fact-checking: FEVER, Climate-FEVER, SciFact Question-Answering: NQ, HotpotQA, FiQA-2018 Bio-Medical IR: TREC-COVID, BioASQ, NFCorpus News Retrieval:

Papers using BEIR (3)

E2Rank: Your Text Embedding can Also be an Effective and Efficient Listwise Reranker2025Large Language Models are Effective Text Rankers with Pairwise Ranking Prompting2023 · 18 citesBeyond Yes and No: Improving Zero-Shot LLM Rankers via Scoring Fine-Grained Relevance Labels2023 · 7 cites