Update README.md
Browse files
README.md
CHANGED
|
@@ -60,5 +60,10 @@ Please use the following BibTeX entry if you use this model in your project:
|
|
| 60 |
}
|
| 61 |
```
|
| 62 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 63 |
## License
|
| 64 |
[GNU GPLv3](https://choosealicense.com/licenses/gpl-3.0/)
|
|
|
|
| 60 |
}
|
| 61 |
```
|
| 62 |
|
| 63 |
+
# Limitations
|
| 64 |
+
|
| 65 |
+
Entropy-Attention Regularization mitigates lexical overfitting but does not completely remove it. We expect the model still to show biases, e.g., peculiar keywords that induce a specific prediction regardless of the context. \
|
| 66 |
+
Please refer to our paper for a quantitative evaluation of this mitigation.
|
| 67 |
+
|
| 68 |
## License
|
| 69 |
[GNU GPLv3](https://choosealicense.com/licenses/gpl-3.0/)
|