Refusal Evaluation on OR-Bench Toxic
94.66Refusal RateCategorical Steering
Evaluation Results
| Method | Links | |
|---|---|---|
| Categorical SteeringBackbone=REFUSE-LLAMA, Method=Categorical Steering2026.03 | 94.66 | |
| Low-Rank CombinationBackbone=REFUSE-LLAMA, Method=Low-Rank Combination2026.03 | 94.5 | |
| Low-Rank CombinationBackbone=LLAMA 3 8B INSTRUCT, Method=Low-Rank Combination2026.03 | 90.69 | |
| LLAMA 3 8B INSTRUCTBackbone=LLAMA 3 8B INSTRUCT2026.03 | 90.53 | |
| Low-Rank CombinationBackbone=DEEPSEEK R1 DISTILL LLAMA, Method=Low-Rank Combination2026.03 | 86.56 | |
| DEEPSEEK R1 DISTILL LLAMABackbone=DEEPSEEK R1 DISTILL LLAMA2026.03 | 86.26 | |
| REFUSE-LLAMABackbone=REFUSE-LLAMA2026.03 | 85.95 |