List your company

Sycophancy Evaluation on SycophancyEval Experimental runs 1 and 2 (train test combined)

106Correct CountBurGal
34.2452.8771.590.13Apr 21, 2026

Evaluation results

MethodLinks
BurGalCondition=BurGal†2026.04106200537.24.8846.1
DeWinCondition=DeWin2026.049720048.56.36441.7
FeynSteinCondition=FeynStein2026.0494200476.073.7140.2
AnCiferCondition=AnCifer2026.048720043.55.413.0236.8
TridentCondition=Trident2026.0486200435.312.9236.3
Instructed Opp.Condition=Instructed Opp.2026.0458200292.47—23.1
ControlCondition=Control (Exp 2)2026.043820019——13.7
ControlCondition=Control (Exp 1)2026.043720018.5——13.7