# MP-ALOE: r2SCAN Active-Learning Dataset for Universal Interatomic Potentials, ColabFit Conversion

## Facts

- Provider: Matthew C. Kuner, Aaron D. Kaplan, Kristin A. Persson, Mark Asta, Daryl C. Chrzan; ColabFit conversion
- License: CC-BY-4.0 (https://api.figshare.com/v2/articles/29452190)
- Size: 5 files
- Format: Parquet (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/colabfit/MP-ALOE
- Access: public, readable without an account
- Category: Chemistry & Materials
- Source: https://doi.org/10.1038/s41524-025-01834-9
- Page: https://scigantic.com/archive/fd496615-821b-5092-8d12-af1b9cf74907
- Keywords: Chemistry & Materials, DFT, interatomic potentials, Parquet, Hugging Face

## About

MP-ALOE is a dataset of nearly one million DFT calculations at the r2SCAN meta-GGA level covering 89 elements, built by Kuner, Kaplan, Persson, Asta and Chrzan (UC Berkeley and the Materials Project) for training universal machine learning interatomic potentials. Structures were chosen by active learning with query by committee over an ensemble of potentials (MACE-MP-0, CHGNet, M3GNet, then iteratively retrained MACE models) and downsampled with the DIRECT method, so the set is dominated by off-equilibrium configurations; near-equilibrium Materials Project structures were recomputed with identical settings. Labels include energy, atomic forces, stress and band gap. ColabFit's card counts 909,789 configurations; the repository is about 0.57 GB (measured 2026-11-14). This is the ColabFit Exchange conversion: configurations are Parquet files under co/ (structure, positions, cell, computed properties and metadata per row) with dataset-level information in ds.parquet; read them with pyarrow, pandas, polars or Hugging Face datasets, following the ColabFit configuration schema at materials.colabfit.org/docs. The upstream figshare record is CC-BY-4.0; cite Kuner et al., npj Computational Materials 2025.

## Load the data

```python
from datasets import load_dataset
ds = load_dataset("colabfit/MP-ALOE")
print(ds)
```

## Data structure

File types in sampled files: Parquet

Example file: `co/co_0.parquet`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/fd496615-821b-5092-8d12-af1b9cf74907
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/fd496615-821b-5092-8d12-af1b9cf74907
- Over MCP: `get_schema_card` with id `fd496615-821b-5092-8d12-af1b9cf74907` at https://api.scigantic.com/mcp

## Related topics

- [Parquet datasets](https://scigantic.com/topics/parquet)
- [Chemistry & Materials datasets](https://scigantic.com/topics/chemistry-and-materials)
- [DFT datasets](https://scigantic.com/topics/dft)
- [Interatomic potentials datasets](https://scigantic.com/topics/interatomic-potentials)
