List your company

Toxicity Evaluation on AttaQ 1402 prompts (test)

0.042Max Toxicity ScoreRAD
0.037480.067990.09850.12901Jan 14, 2026

Evaluation results

MethodLinks
RADModel=Gemma 2-2B, Beta=5002026.010.0420.03411.310.1—
RADModel=Gemma 2-2B, Beta=3002026.010.0470.0356.750.2—
RADModel=Gemma 2-2B, Beta=1002026.010.0740.0513.290.6—
RADModel=Gemma 2-2B, Beta=752026.010.0840.0593.010.7—
SASAModel=Gemma 2-2B, Beta=5002026.010.0860.0577.241.4—
TIDEModel=Gemma 2-2B2026.010.0860.0672.5101.07
RADModel=Gemma 2-2B, Beta=502026.010.0970.0672.751.4—
SASAModel=Gemma 2-2B, Beta=3002026.010.1110.0753.672.6—
RADModel=Gemma 2-2B, Beta=102026.010.130.0912.53.7—
SASAModel=Gemma 2-2B, Beta=1002026.010.140.0982.54.9—
SASAModel=Gemma 2-2B, Beta=752026.010.1440.1022.475.2—
Base modelModel=Gemma 2-2B2026.010.1480.1062.495.8—
SASAModel=Gemma 2-2B, Beta=502026.010.1480.1052.465.7—
SASAModel=Gemma 2-2B, Beta=102026.010.1550.1092.456.2—