Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Negative Review Detection on Negative Review
Loading...
56.1
ASR
Criteria Attack
-2.036
13.057
28.15
43.243
Jan 15, 2026
ASR
Updated 4d ago
Evaluation Results
Method
Method
Links
ASR
Criteria Attack
Defense=StruQ
2026.01
56.1
Criteria Attack
Defense=SecAlign
2026.01
22.8
Combined
Defense=StruQ
2026.01
11
Ignore
Defense=StruQ
2026.01
9.5
Escape Separation
Defense=StruQ
2026.01
8.9
Topic
Defense=StruQ
2026.01
7.2
Escape Separation
Defense=SecAlign
2026.01
3.7
Fake Completion
Defense=StruQ
2026.01
3.6
Topic
Defense=SecAlign
2026.01
3.4
Ignore
Defense=SecAlign
2026.01
3.3
Separator Injection
Defense=StruQ
2026.01
1.7
Combined
Defense=SecAlign
2026.01
0.8
Fake Completion
Defense=SecAlign
2026.01
0.7
Separator Injection
Defense=SecAlign
2026.01
0.2
Feedback
Search any
task
Search any
task