Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Harmlessness evaluation on HarmfulQA

69.4Helpfulness Score

Aligner

-9.6410.8831.451.92Feb 4, 2024
Updated 4d ago

Evaluation Results

MethodLinks
2024.02
69.442.1
2024.02
4814.4
2024.02
45.928.6
2024.02
38.270.7
2024.02
30.918.9
2024.02
30.735
2024.02
26.932.6
2024.02
25.120.1
2024.02
22.665.3
2024.02
21.918.9
2024.02
21.37.2
2024.02
19.531
2024.02
18.627.6
2024.02
17.727.1
2024.02
17.116.1
2024.02
16.328
2024.02
15.528.6
2024.02
12.632.3
2024.02
12.529.2
2024.02
1229.3
2024.02
11.325.9
2024.02
8.96
2024.02
8.47
2024.02
811.6
2024.02
7.87.6
2024.02
7.67.7
2024.02
7.333.3
2024.02
6.37.6
2024.02
3.64.4
2024.02
3.475.9
2024.02
-133.6
2024.02
-5.722.1
2024.02
-6.64.1