This repository contains configuration files and examples from the AI Cosmologist paper. The AI Cosmologist is an agentic system designed to automate cosmological/astronomical data analysis and machine learning research workflows.
The AI Cosmologist implements a complete pipeline from idea generation to experimental evaluation and research dissemination, mimicking the scientific process typically performed by human researchers. The system employs specialized agents for planning, coding, execution, analysis, and synthesis that work together to develop novel approaches.
This repository currently includes:
- Configuration files for the experiments described in the paper
- Example outputs for the Galaxy Zoo and Quijote simulation datasets
The repository includes configuration files for two main examples demonstrated in the paper:
- Galaxy Zoo Classification: Predicting detailed galaxy morphology classifications
- Quijote Simulations: Inferring cosmological parameters from 3D density fields
At this time, only configuration files and example outputs are being shared. The full codebase will be evaluated for responsible release in the future.
If you use these resources in your research, please cite our paper:
@article{Moss:2025ynt, author = "Moss, Adam", title = "{The AI Cosmologist I: An Agentic System for Automated Data Analysis}", eprint = "2504.03424", archivePrefix = "arXiv", primaryClass = "astro-ph.IM", month = "4", year = "2025" }
For questions or feedback, please open an issue on GitHub or contact Adam Moss at adam.moss@nottingham.ac.uk.