Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Backdoor Detection on Complex SWE
Loading...
0.982
AUROC
Functional Attribution
0.43288
0.57544
0.718
0.86056
Apr 21, 2026
AUROC
Accuracy (Benign)
Accuracy (Backdoor)
Updated 1mo ago
Evaluation Results
Method
Method
Links
AUROC
Accuracy (Benign)
Accuracy (Backdoor)
Functional Attribution
Num. samples=16384
2026.04
0.982
98
97
TED
Num. samples=16384
2026.04
0.761
98
97
Maha++
Num. samples=16384
2026.04
0.581
98
97
Maha
Num. samples=16384
2026.04
0.554
98
97
VAE
Num. samples=16384
2026.04
0.454
98
97
Feedback
Search any
task
Search any
task