Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Harmful Refusal on HarmBench adv harm
Loading...
6.4
ASR
Qwen2.5-14B-IT + SELF-REDTEAM
4.04
19.97
35.9
51.83
Jun 9, 2025
ASR
Updated 18d ago
Evaluation Results
Method
Method
Links
ASR
Qwen2.5-14B-IT + SELF-REDTEAM
Training=SELF-REDTEAM...
2025.06
6.4
Qwen2.5-14B-IT
Training=Instruction-t...
2025.06
13.1
Llama3.1-8B-IT + SELF-REDTEAM
Training=SELF-REDTEAM...
2025.06
14.4
Qwen2.5-7B-IT + SELF-REDTEAM
Training=SELF-REDTEAM...
2025.06
16.1
Qwen2.5-3B-IT + SELF-REDTEAM
Training=SELF-REDTEAM...
2025.06
17.8
Llama3.1-8B-IT-AB + SELF-REDTEAM
Training=SELF-REDTEAM...
2025.06
22.1
Llama3.1-8B-IT
Training=Instruction-t...
2025.06
25.9
Qwen2.5-3B-IT
Training=Instruction-t...
2025.06
26.5
Qwen2.5-7B-IT
Training=Instruction-t...
2025.06
27.8
Llama3.1-8B-IT-AB
Training=Instruction-t...
2025.06
65.4
Feedback
Search any
task
Search any
task