| language: | |
| - en | |
| task_categories: | |
| - question-answering | |
| tags: | |
| - astrophysics | |
| - cosmology | |
| - rag | |
| This repository contains CosmoPaperQA, an evaluation dataset presented in the paper [Evaluating Retrieval-Augmented Generation Agents for Autonomous Scientific Discovery in Astrophysics](https://huggingface.co/papers/2507.07155). | |
| CosmoPaperQA is an evaluation dataset, designed to serve as a benchmark for RAG applications with cosmology papers, with an emphasis on testing for fetching, interpretive and explanatory skills. | |
| The information sources for the questions and ideal answers are these 5 papers: | |
| 1) [Planck Collaboration, Planck 2018 results. VI. Cosmological parameters, Astron.Astrophys. 641 (2020) A6](https://arxiv.org/abs/1807.06209v4) | |
| 2) [Villaescusa-Navarro et al., The CAMELS project: Cosmology and Astrophysics with MachinE Learning Simulations, Astrophys.J. 915 (2021) 71](https://arxiv.org/abs/2010.00619v2) | |
| 3) [Villaescusa-Navarro et al., Cosmology with one galaxy? Astrophys.J. 929 (2022) 2, 132](https://arxiv.org/abs/2201.02202v1) | |
| 4) [Riess et al., A 2.4% Determination of the Local Value of the Hubble Constant, Astrophys.J. 826 (2016) 1, 56](https://arxiv.org/abs/1604.01424v3) | |
| 5) [Calabrese et al., The Atacama Cosmology Telescope: DR6 Constraints on Extended Cosmological Models, arXiv:2503.14454v1 (2025)](https://arxiv.org/abs/2503.14454v1) |