Openethos/ethos-transparency-classifier
Text Classification
β’
Updated
β’
12
organization_name
stringclasses 4
values | mission_focus
stringclasses 4
values | year_founded
int64 2.02k
2.03k
| transparency_score
int64 7
10
|
|---|---|---|---|
OpenEthos Research Collective
|
Explainable AI & Ethics
| 2,025
| 10
|
OpenAI
|
General Intelligence
| 2,015
| 7
|
Anthropic
|
AI Safety
| 2,021
| 8
|
Hugging Face
|
Open Source Community
| 2,016
| 9
|
This dataset contains curated text samples regarding AI governance and ethical manifestos. It is designed to help researchers train models that can identify transparency commitments in technical documentation.
As part of the OpenEthos Research Collective, this dataset has been audited for:
The data is provided in CSV format with the following columns:
source_url: The origin of the text.text_content: The raw manifesto or report text.ethics_label: (e.g., 'Transparent', 'Opaque', 'Neutral').