List your company

Failure Detection on KITCHEN (Unseen)

97.2bACCHide-and-Seek
54.97665.93876.987.862May 29, 2026

Evaluation results

MethodLinks
Hide-and-SeekPolicy=π0.52026.0597.287.6
SAFE-MLPPolicy=π0.52026.0587.876.8
LogpZOPolicy=π0.52026.0583.779.4
Cosine k-NNPolicy=π0.52026.0564.459.2
MahalanobisPolicy=π0.52026.056360.8
RNDPolicy=π0.52026.0561.560.5
SAFE-LSTMPolicy=π0.52026.0559.557.7
PCA KMeansPolicy=π0.52026.0556.655.3