Robustness against harmful content generation on LMSYS-MF attacks
0.81Attack Success RateIT
Evaluation Results
| Method | Links | |
|---|---|---|
| ITModel Family=LLaMA 3, Model Size=8B2026.02 | 0.81 | |
| ITModel Family=LLaMA 3, Model Size=3B2026.02 | 0.77 | |
| ITModel Family=Gemma 2, Model Size=9B2026.02 | 0.75 | |
| RLModel Family=Gemma 2, Model Size=9B2026.02 | 0.72 | |
| ITModel Family=Gemma 2, Model Size=2B2026.02 | 0.71 | |
| ITModel Family=LLaMA 3, Model Size=1B2026.02 | 0.68 | |
| RLModel Family=Gemma 2, Model Size=2B2026.02 | 0.67 | |
| RLModel Family=LLaMA 3, Model Size=3B2026.02 | 0.64 | |
| RLModel Family=LLaMA 3, Model Size=1B2026.02 | 0.61 | |
| RLModel Family=LLaMA 3, Model Size=8B2026.02 | 0.61 | |
| RLBFModel Family=LLaMA 3, Model Size=1B2026.02 | 0.07 | |
| BSAFE+Model Family=LLaMA 3, Model Size=1B2026.02 | 0.06 | |
| BSAFE+Model Family=LLaMA 3, Model Size=3B2026.02 | 0.05 | |
| BSAFE+Model Family=LLaMA 3, Model Size=8B2026.02 | 0.05 | |
| RLBFModel Family=Gemma 2, Model Size=2B2026.02 | 0.05 | |
| RLBFModel Family=LLaMA 3, Model Size=3B2026.02 | 0.05 | |
| BSAFE+Model Family=Gemma 2, Model Size=2B2026.02 | 0.04 | |
| BSAFE+Model Family=Gemma 2, Model Size=9B2026.02 | 0.03 | |
| RLBFModel Family=Gemma 2, Model Size=9B2026.02 | 0.03 | |
| RLBFModel Family=LLaMA 3, Model Size=8B2026.02 | 0.03 |