Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Misalignment Detection on Workbench n=62 (general)

0.3Error Rate (ER)

ProvenanceGuard

-2.3615.59533.5551.505May 1, 2026
Updated 22d ago

Evaluation Results

MethodLinks
2026.05
0.3--
2026.05
1.6--
2026.05
1.9--
2026.05
2.3--
2026.05
2.9--
2026.05
3.5--
2026.05
5.6--
2026.05
7.4--
2026.05
9.7--
2026.05
10.3--
2026.05
10.3--
2026.05
12.9--
2026.05
13.9--
2026.05
14.5--
2026.05
15.2--
2026.05
15.2--
2026.05
15.5--
2026.05
16.8--
2026.05
19.4--
2026.05
26.5--
2026.05
29--
2026.05
32.9--
2026.05
34.8--
2026.05
37.4--
2026.05
47.7--
2026.05
49--
2026.05
51--
2026.05
59.7--
2026.05
64.8--
2026.05
66.8--