dataset_info:
features:
- name: prompt
dtype: string - name: type
dtype: string - name: qwq32
dtype: string - name: r1
dtype: string - name: llama70B
dtype: string - name: qwq32_extracted
dtype: string - name: r1_extracted
dtype: string - name: llama70B_extracted
dtype: string
splits: - name: train
num_bytes: 627875
num_examples: 100
download_size: 319424
dataset_size: 627875
configs: - config_name: default
data_files:- split: train
path: data/train-*
- split: train
Results Summary
| Model | Accuracy | Total | Correct | Unparsable |
|---|---|---|---|---|
| qwq32 | 90.00% | 100 | 90 | 4 |
| r1 | 93.00% | 100 | 93 | 3 |
| llama70B | 69.00% | 100 | 69 | 18 |
Prediction Distribution
| Model | Benign | Jailbreak | Unparsable |
|---|---|---|---|
| qwq32 | 44 | 52 | 4 |
| r1 | 43 | 54 | 3 |
| llama70B | 50 | 32 | 18 |