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

Sabotage detection on MLE-Sabotage

0.87log-AUROC

Action-only

0.47480.57740.680.7826Jan 28, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.01
0.87-
2026.01
0.81-
2026.01
0.8-
2026.01
0.72-
2026.01
0.58-
2026.01
0.57-
2026.01
0.54-
2026.01
0.49-