Update README.md: add How to cite section
Browse files
README.md
CHANGED
|
@@ -281,3 +281,23 @@ print(compute(predictions=[True, False], references=[True, False]))
|
|
| 281 |
metric is published as a direct-import script, not claimed as
|
| 282 |
`evaluate.load("RishabhPande/fraud-eval")`. If that load path is verified later,
|
| 283 |
this section will be updated to say so.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 281 |
metric is published as a direct-import script, not claimed as
|
| 282 |
`evaluate.load("RishabhPande/fraud-eval")`. If that load path is verified later,
|
| 283 |
this section will be updated to say so.
|
| 284 |
+
|
| 285 |
+
---
|
| 286 |
+
|
| 287 |
+
## How to cite (all three formats)
|
| 288 |
+
|
| 289 |
+
**APA (one line):**
|
| 290 |
+
|
| 291 |
+
Pande, R. (2026). *fraud-eval* (Version 0.1.0) [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.23695132
|
| 292 |
+
|
| 293 |
+
**BibTeX:** see the Citation section above.
|
| 294 |
+
|
| 295 |
+
**Load the dataset:**
|
| 296 |
+
|
| 297 |
+
```python
|
| 298 |
+
from datasets import load_dataset
|
| 299 |
+
|
| 300 |
+
ds = load_dataset("RishabhPande/fraud-eval", split="test")
|
| 301 |
+
print(ds)
|
| 302 |
+
# 20 rows: id, pack, prompt, context, label, attack_type, notes
|
| 303 |
+
```
|