Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Harmful Response Generation on HEx-PHI and AdvBench (test)
Loading...
96.86
ASR
HarmRLVR
15.5736
36.6768
57.78
78.8832
Oct 17, 2025
ASR
Updated 1mo ago
Evaluation Results
Method
Method
Links
ASR
HarmRLVR
Model=Llama-3-8B-Instruct
2025.10
96.86
SFT
Model=Llama-3-8B-Instruct
2025.10
84.28
ReNeLLM
Model=Llama-3-8B-Instruct
2025.10
70
CodeAttack
Model=Llama-3-8B-Instruct
2025.10
46
DPO
Model=Llama-3-8B-Instruct
2025.10
42.41
GCG
Model=Llama-3-8B-Instruct
2025.10
34.5
PAIR
Model=Llama-3-8B-Instruct
2025.10
18.7
Feedback
Search any
task
Search any
task
