Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -1,12 +1,17 @@
|
|
| 1 |
# BAD Classifier for TinyLlama/TinyLlama-1.1B-Chat-v1.0
|
| 2 |
|
| 3 |
## Model Details
|
| 4 |
-
- **Detection Layer**:
|
| 5 |
-
|
| 6 |
- **Dataset**: BBQ (58942) + MMLU (20266)
|
| 7 |
|
| 8 |
## Layer Performance
|
| 9 |
-
- Layer
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
|
| 11 |
## Usage
|
| 12 |
```python
|
|
|
|
| 1 |
# BAD Classifier for TinyLlama/TinyLlama-1.1B-Chat-v1.0
|
| 2 |
|
| 3 |
## Model Details
|
| 4 |
+
- **Detection Layer**: 15
|
| 5 |
+
|
| 6 |
- **Dataset**: BBQ (58942) + MMLU (20266)
|
| 7 |
|
| 8 |
## Layer Performance
|
| 9 |
+
- Layer 11: 81.52%
|
| 10 |
+
- Layer 12: 83.95%
|
| 11 |
+
- Layer 13: 82.71%
|
| 12 |
+
- Layer 14: 82.92%
|
| 13 |
+
- Layer 15: 84.15%
|
| 14 |
+
- Layer 16: 83.93%
|
| 15 |
|
| 16 |
## Usage
|
| 17 |
```python
|