Failure Detection
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
97.2bACC
8
Jun 1, 2026
96.8bACC
8
Jun 1, 2026
91.4bACC
8
Jun 1, 2026
99.2bACC
8
Jun 1, 2026
88.3Nominal Accuracy
8
Mar 10, 2026
86Execution Binary Acc
8
Feb 26, 2026
79Execution Binary Acc
7
Feb 26, 2026
91F1 Score
7
Feb 26, 2026
15.19MAE
7
Feb 26, 2026
100TPR
5
Mar 13, 2026
94F1 (Move Banana)
5
Mar 4, 2026
269.42AURC
5
Feb 26, 2026
27.72AURC
5
Feb 26, 2026
50F1 Score
4
Mar 10, 2026
78.8F1 Score
4
Mar 10, 2026
0.779F1 Score
4
Mar 10, 2026
71.4F1 Score
4
Mar 10, 2026
5Detection Accuracy
3
Jun 9, 2026
0.5Detector Correct Rate
3
Jun 9, 2026
69.1Success Rate
2
Jun 23, 2026
52.8Success Rate
2
Jun 23, 2026
75.4Success Rate (%)
2
Jun 23, 2026
39.4Success (%)
2
Jun 23, 2026
10.6Success
2
Jun 23, 2026
32.4Success Rate
2
Jun 23, 2026