Datasets:

Modalities:
Text
Formats:
csv
Languages:
English
Size:
< 1K
ArXiv:
Libraries:
Datasets
pandas
License:
fmti-indicators / README.md
mariagrandury's picture
update dataset description
231a0b4
|
raw
history blame
No virus
1.77 kB
---
license: cc-by-4.0
language:
- en
pretty_name: The Foundation Model Transparency Index
size_categories:
- n<1K
---
# Dataset Card for Dataset Name
The Foundation Model Transparency Index is an ongoing initiative of the CRFM to comprehensively assess the transparency of foundation model developers.
- **Created by:** Center for Research on Foundation Models
- **License:** [Creative Commons Attribution 4.0 International](https://github.com/stanford-crfm/fmti/blob/main/LICENSE.md)
## Dataset Sources
- **Repository:** https://github.com/stanford-crfm/fmti
- **Paper:** https://arxiv.org/abs/2310.12941
- **Website:** https://crfm.stanford.edu/fmti/
## Uses
Assess the transparency of foundation model developers.
## Dataset Structure
- `Domain`: `Upstream` (model building), `Model` (properties and function), and `Downstream` (distribution and usage).
- `Subdomain`: Data, labor, compute, methods, model basics, model access, capabilities, risks, mitigations, distribution, usage policy, feedback, and impact.
- `Indicator`: Name of the indicator.
- `Definition`: Question for the developers of the model.
- `Notes`: Explanation of how to receive the point.
- References: `Reference_1` and `Reference_2` with their corresponding `Link_1` and `Link_2`.
## Citation
```
@article{bommasani2023fmti,
author = {Bommasani, Rishi and
Klyman, Kevin and
Longpre, Shayne and
Kapoor, Sayash and
Maslej, Nestor and
Xiong, Betty and
Zhang, Daniel and
Liang, Percy},
title = {The Foundation Model Transparency Index},
month = oct,
year = 2023,
url = {https://crfm.stanford.edu/fmti/FMTI.pdf}
}
```