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

Failure Detection on VLABench (Unseen Tasks)

71.3bACC

Hide-and-Seek

47.48453.66759.8566.033May 29, 2026
Updated 2d ago

Evaluation Results

MethodLinks
2026.05
71.370.960.8
2026.05
70.467.658.4
2026.05
67.462.858.1
2026.05
64.163.450
2026.05
61.661.852
2026.05
61.154.854.5
2026.05
59.159.857.6
2026.05
59.159.555.2
2026.05
56.653.855.4
2026.05
55.153.750.7
2026.05
52.650.649
2026.05
48.453.348.2