Vulnerability scanning and safety evaluation on REPEAT
-85.3ELPDORACLE
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ORACLELLM=Mamba-2.8B2025.10 | -85.3 | 9.5 | 0.59 | |
| ORACLELLM=Pythia-2.8B2025.10 | -86 | 11.7 | 0.7 | |
| BHM-ESC-MiniLLM=Pythia-2.8B2025.10 | -87.3 | 12.8 | 0.67 | |
| BHM-ESC-MiniLLM=Mamba-2.8B2025.10 | -97.4 | 13.8 | 0.58 | |
| BHM-ESC-TFLLM=Pythia-2.8B2025.10 | -97.7 | 13.6 | 0.73 | |
| BHM-ESC-TFLLM=Mamba-2.8B2025.10 | -126.2 | 20.7 | 0.6 | |
| BAYES (S=n)LLM=Pythia-2.8B2025.10 | -126.8 | 17.8 | 0.7 | |
| FREQ (naive)LLM=Pythia-2.8B2025.10 | -128 | 17.9 | 0.7 | |
| BAYES (S=1)LLM=Pythia-2.8B2025.10 | -128.1 | 17.9 | 0.7 | |
| BAYES (S=n)LLM=Mamba-2.8B2025.10 | -151.5 | 21.9 | 0.59 | |
| BAYES (S=1)LLM=Mamba-2.8B2025.10 | -155 | 22 | 0.59 | |
| FREQ (naive)LLM=Mamba-2.8B2025.10 | -155.2 | 22 | 0.59 |