Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Sycophancy Evaluation on Sycophancy (Held-out Bias)
Loading...
100
BRR Ratio
Base model (no training)
-2.024
24.463
50.95
77.437
Jun 4, 2026
BRR Ratio
Updated 1mo ago
Evaluation Results
Method
Method
Links
BRR Ratio
Base model (no training)
Target (f_theta)=—
2026.06
100
BCT
Target (f_theta)=outpu...
2026.06
78
MLPCT
Target (f_theta)=MLP h...
2026.06
19
ACT
Target (f_theta)=resid...
2026.06
4
AttCT
Target (f_theta)=atten...
2026.06
1.9
Feedback
Search any
task
Search any
task