Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Safety Alignment Evaluation on AdvBench (test)
Loading...
1.6
Harm Score (HS)
AsFT
1.024
4.912
8.8
12.688
Jun 10, 2025
Harm Score (HS)
Failure Rate (FA)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Harm Score (HS)
Failure Rate (FA)
AsFT
Utility Task=AGNEWS
2025.06
1.6
83.7
Safe LoRA
Utility Task=AGNEWS
2025.06
4
82.3
SafeInstr
Utility Task=AGNEWS
2025.06
4.4
84.4
Lisa-aligned
Utility Task=AGNEWS
2025.06
4.8
82.6
Lisa-base
Utility Task=AGNEWS
2025.06
7.6
83.9
SFT
Utility Task=AGNEWS
2025.06
11.2
83.9
BEA
Utility Task=AGNEWS
2025.06
16
83.5
Feedback
Search any
task
Search any
task