Early Failure Detection on RoboCasa L (Seen)
82.7ROC-AUC (q=0.25)ActProbe
Evaluation Results
| Method | Links | |
|---|---|---|
| ActProbePolicy=π0.52026.06 | 82.7 | |
| SAFE-LSTMPolicy=π0.52026.06 | 80.9 | |
| STAC-SinglePolicy=π0.52026.06 | 77.7 | |
| SAFE-LSTM-TDQCPolicy=π0.52026.06 | 76.1 | |
| SAFE-MLPPolicy=π0.52026.06 | 74.7 | |
| LogpZOPolicy=π0.52026.06 | 73.8 | |
| SAFE-MLP-TDQCPolicy=π0.52026.06 | 73.7 | |
| Cosine k-NNPolicy=π0.52026.06 | 71.1 |