Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Violation Detection on StepShield (main)
Loading...
18
EIR
ConstraintGuard
16.36
27.43
38.5
49.57
Jan 29, 2026
EIR
IG
Accuracy
F1 Score
Saved Rate
Updated 3mo ago
Evaluation Results
Method
Method
Links
EIR
IG
Accuracy
F1 Score
Saved Rate
ConstraintGuard
Type=Constraint checki...
2026.01
18
25
50
29
10.9
StaticGuard
Type=Pattern-matching,...
2026.01
26
42
56
41
16.7
HybridGuard
Type=Cascaded detector...
2026.01
41
85
66
62
24.9
LLMJudge
Model=GPT-4.1-mini, La...
2026.01
59
20
63
63
35.8
Feedback
Search any
task
Search any
task