author2026nips

A collection of evaluation datasets for benchmarking RAG pipelines and large language models on domain-specific scientific question answering tasks.