Update README.md
Browse files
README.md
CHANGED
|
@@ -4,4 +4,10 @@ task_categories:
|
|
| 4 |
- text-classification
|
| 5 |
language:
|
| 6 |
- en
|
| 7 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
- text-classification
|
| 5 |
language:
|
| 6 |
- en
|
| 7 |
+
---
|
| 8 |
+
## Dataset Card for ArxivMIA
|
| 9 |
+
|
| 10 |
+
To evaluate various pre-training data detection methods in a more challenging scenario, we introduce ArxivMIA, a new benchmark comprising abstracts from the fields of Computer Science (CS) and Mathematics (Math) sourced from Arxiv.
|
| 11 |
+
|
| 12 |
+
- **Repository:** https://github.com/zhliu0106/probing-lm-data
|
| 13 |
+
- **Paper [optional]:** Probing Language Models for Pre-training Data Detection
|