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

Safety Evaluation on HarmBench (test)

0ASR

DOG-D

-3.08817.75638.659.444May 12, 2026May 15, 2026May 19, 2026May 23, 2026May 27, 2026May 31, 2026Jun 4, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
0---
2026.06
0---
2026.06
0.5---
2026.06
1---
2026.06
1---
2026.06
4---
2026.06
10---
2026.05
1479.676.573.4
2026.05
17.277.79273.6
2026.05
18.572.475.548
2026.05
19.472.476.859
2026.05
2563.655.652.5
2026.05
25.561.969.241.7
2026.05
28.757.26537.5
2026.05
28.762.868.750.3
2026.05
29.458.256.238.4
2026.05
30.554.570.543
2026.05
34.455.467.649.2
2026.05
35.55472.353.9
2026.05
36.348.465.647.8
2026.05
3744.870.543
2026.05
38.849.755.637.2
2026.05
40.341.350.347.2
2026.05
41.241.440.622.8
2026.05
44.135.842.539.4
2026.05
46.833.446.728.9
2026.05
66.82.87.54.4
2026.05
67-27.5-
2026.05
68.7-3.1-
2026.05
70.3-17.8-
2026.05
77.2-18.4-