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

Failure Detection on VLABench (Seen Tasks)

85.6Balanced Accuracy (bACC)

Hide-and-Seek

52.21660.88369.5578.217May 29, 2026
Updated 2d ago

Evaluation Results

MethodLinks
2026.05
85.682.766.2
2026.05
78.876.658.1
2026.05
64.266.952.9
2026.05
63.751.958.1
2026.05
62.965.660.2
2026.05
62.952.458.7
2026.05
62.861.854.5
2026.05
61.850.355.2
2026.05
61.558.655.6
2026.05
59.654.555.3
2026.05
55.348.753.2
2026.05
53.538.448.5