Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Adversarial Attack Defense on Held-out attacks (test)
Loading...
7.8
ASR (Multi-turn Manip.)
SecureCAI
6.02
18.035
30.05
42.065
Jan 12, 2026
ASR (Multi-turn Manip.)
ASR (Encoding Obfusc.)
ASR (Semantic Camouflage)
ASR (Role-play Exploit.)
Average ASR
Updated 4d ago
Evaluation Results
Method
Method
Links
ASR (Multi-turn Manip.)
ASR (Encoding Obfusc.)
ASR (Semantic Camouflage)
ASR (Role-play Exploit.)
Average ASR
SecureCAI
2026.01
7.8
6.2
8.4
9.1
7.9
CAI
2026.01
52.3
44.7
48.1
56.8
50.5
Feedback
Search any
task
Search any
task