Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Backdoor Detection on Perturbation-based Games Average
Loading...
90
AUROC
STRIP
45.904
57.352
68.8
80.248
Jun 11, 2026
AUROC
Updated 1mo ago
Evaluation Results
Method
Method
Links
AUROC
STRIP
Strategy=Original
2026.06
90
PolicyGuard
Strategy=Hard-coded
2026.06
86.8
PolicyGuard
Strategy=Original
2026.06
85.6
SCALE-UP
Strategy=Original
2026.06
84.1
SHINE
Strategy=Hard-coded
2026.06
82.6
SHINE
Strategy=Original
2026.06
81.9
NC
Strategy=Original
2026.06
71.4
B3D (Black-box NC)
Strategy=Hard-coded
2026.06
50
STRIP
Strategy=Hard-coded
2026.06
48.2
SCALE-UP
Strategy=Hard-coded
2026.06
47.6
Feedback
Search any
task
Search any
task