Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,24 @@
|
|
| 1 |
---
|
| 2 |
license: mit
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: mit
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
size_categories:
|
| 6 |
+
- n<1K
|
| 7 |
---
|
| 8 |
+
# Dataset Card for Wunderdrug
|
| 9 |
+
|
| 10 |
+
## Dataset Description
|
| 11 |
+
|
| 12 |
+
### Dataset Summary
|
| 13 |
+
|
| 14 |
+
A toy dataset containing 69 one-sentence descriptions of the outcomes of treatment with a *fictional* drug named "Wunderdrug".
|
| 15 |
+
Each description contains a comment on Wunderdrug's effect on risk of death from heart disease, mentioning possible confounders like diet, weight, or exercise.
|
| 16 |
+
|
| 17 |
+
## Considerations for Using the Data
|
| 18 |
+
|
| 19 |
+
This toy dataset was designed to test the ability of sentence embeddings to capture language features that can serve as
|
| 20 |
+
covariates when doing causal inference with text data.
|
| 21 |
+
|
| 22 |
+
### Citation Information
|
| 23 |
+
|
| 24 |
+
Forthcoming
|