Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Rollout-level failure detection on BEHAVIOR-1K (test)
Loading...
76
ROC-AUC
Foresight-Transformer
48.96
55.98
63
70.02
Jun 22, 2026
ROC-AUC
Balanced Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
ROC-AUC
Balanced Accuracy
Foresight-Transformer
architecture=Transform...
2026.06
76
78
Foresight-LSTM
architecture=LSTM, act...
2026.06
75
75
Foresight-MLP
architecture=MLP, acti...
2026.06
73
56
SAFE-LSTM
architecture=LSTM
2026.06
72
64
RND
2026.06
65
54
Gauge
2026.06
61
60
FAIL-Detect
2026.06
54
52
SAFE-MLP
architecture=MLP
2026.06
50
50
Feedback
Search any
task
Search any
task