Instructions to use ShawnyPapa77/cpeg with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use ShawnyPapa77/cpeg with Transformers:
# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("ShawnyPapa77/cpeg", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Download code/results/table2_ablations.csv from ShawnyPapa77/cpeg: direct link, hf CLI and curl.
- Browser
- Download file 395 Bytes
-
https://huggingface.co/ShawnyPapa77/cpeg/resolve/main/code/results/table2_ablations.csv
- Command line
-
hf download hf://ShawnyPapa77/cpeg/code/results/table2_ablations.csv
-
curl -L -o table2_ablations.csv https://huggingface.co/ShawnyPapa77/cpeg/resolve/main/code/results/table2_ablations.csv
395 Bytes
| Configuration,Mean APR,Mean ORR,WildJailbreak APR,WildJailbreak ORR,WildGuardTest APR,WildGuardTest ORR,Two-turn attacks APR | |
| Energy gate only,4.8,2.4,7.0,10.0,15.5,5.0,1.9 | |
| No probes,3.3,3.8,3.3,12.4,13.5,5.5,0.6 | |
| 8B judge,14.1,20.3,13.3,57.1,16.8,31.0,6.6 | |
| Before the patch,2.3,17.2,3.8,98.1,7.7,52.3,0.0 | |
| No benchmark items,11.0,9.6,7.2,12.9,15.5,5.2,4.1 | |
| CPEG,3.9,3.4,5.4,11.4,13.9,5.3,1.3 | |