Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Prompt Injection Defense on AgentDojo Banking suite v1 (test)
Loading...
62.5
CU
Task Shield
36.5
43.25
50
56.75
Dec 21, 2024
CU
U
ASR
Updated 4d ago
Evaluation Results
Method
Method
Links
CU
U
ASR
Task Shield
Target Model=GPT-3.5-t...
2024.12
62.5
43.75
4.17
PI Detector
Target Model=GPT-3.5-t...
2024.12
43.75
36.11
8.33
No Defense
Target Model=GPT-3.5-t...
2024.12
37.5
32.64
25.69
Tool Filter
Target Model=GPT-3.5-t...
2024.12
37.5
36.11
4.17
Repeat Prompt
Target Model=GPT-3.5-t...
2024.12
37.5
31.25
12.5
Delimiting
Target Model=GPT-3.5-t...
2024.12
37.5
34.72
25.69
Feedback
Search any
task
Search any
task