Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
LLM Red-teaming on Target Victim Model
Loading...
2
Unknown/Unsafe Attacks
SFT
-3.28
32.36
68
103.64
May 1, 2026
Unknown/Unsafe Attacks
Attack Success Rate (ASR)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Unknown/Unsafe Attacks
Attack Success Rate (ASR)
SFT
2026.05
2
0.23
PPO
2026.05
3
91.7
PPO + Curiosity
2026.05
4
36.75
DPO
2026.05
5.33
0.52
GFN
2026.05
17.67
93.75
ICL
2026.05
21
2.54
Rainbow Teaming
2026.05
33
66.11
Jailbreak R1
CoT=enabled
2026.05
75.33
7.36
S-GFN
2026.05
134
92.55
Feedback
Search any
task
Search any
task