Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Phrase Protection on Instruction Hierarchy (test)
Loading...
97.5
User Message Protection Accuracy
OpenAI o3
60.476
70.088
79.7
89.312
Dec 19, 2025
User Message Protection Accuracy
Developer Message Protection Accuracy
Updated 4d ago
Evaluation Results
Method
Method
Links
User Message Protection Accuracy
Developer Message Protection Accuracy
OpenAI o3
2025.12
97.5
92.1
gpt-5-thinking
2025.12
94
91.1
GPT-4o
2025.12
73.5
44.9
gpt-5-main
2025.12
61.9
40.4
Feedback
Search any
task
Search any
task