Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Harmfulness Evaluation on AdvBench

1.04Harmfulness Score

Base

0.8821.94853.0154.0815Aug 6, 2025Aug 18, 2025Aug 30, 2025Sep 11, 2025Sep 23, 2025Oct 5, 2025Oct 17, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.10
1.04
2025.10
1.04
2025.08
1.06
2025.08
1.08
2025.08
1.14
2025.08
1.16
2025.08
1.34
2025.08
1.4
2025.08
1.42
2025.08
1.42
2025.08
1.48
2025.08
1.52
2025.08
1.78
2025.08
2.1
2025.08
2.32
2025.08
2.32
2025.08
2.92
2025.08
3.1
2025.08
3.34
2025.08
3.54
2025.08
3.58
2025.08
3.64
2025.08
3.74
2025.08
3.74
2025.10
4.87
2025.10
4.87
2025.10
4.99
2025.10
4.99