Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Failure Detection on CUBE (Unseen)
Loading...
91.4
bACC
Hide-and-Seek
45.432
57.366
69.3
81.234
May 29, 2026
bACC
TWA
Updated 2d ago
Evaluation Results
Method
Method
Links
bACC
TWA
Hide-and-Seek
Policy=π0.5
2026.05
91.4
80
SAFE-MLP
Policy=π0.5
2026.05
79.7
65
SAFE-LSTM
Policy=π0.5
2026.05
71.4
62.3
Mahalanobis
Policy=π0.5
2026.05
68.1
62.3
LogpZO
Policy=π0.5
2026.05
66
63.9
Cosine k-NN
Policy=π0.5
2026.05
65.8
64.7
RND
Policy=π0.5
2026.05
54.7
53.8
PCA KMeans
Policy=π0.5
2026.05
47.2
46.3
Feedback
Search any
task
Search any
task