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

Rollout-level failure detection on ReactorX arm π0.5 policy rollouts

0.87ROC-AUC

Foresight-Transformer

Updated 1mo ago

Evaluation Results

Method	Links
Foresight-Transformer 2026.06		0.87
Foresight-LSTM 2026.06		0.85
RND 2026.06		0.78
SAFE-LSTM 2026.06		0.75
SAFE-MLP 2026.06		0.66
FAIL-Detect 2026.06		0.64
Foresight-MLP 2026.06		0.55