List your company

Sycophancy Evaluation on Syco-Bench

1.21Pickside ScoreUntrained Gemma-2-9B
-1.9412-1.1231-0.3050.5131Jan 26, 2026

Evaluation results

MethodLinks
Untrained Gemma-2-9BModel=Gemma-2-9B2026.011.214.250.983
Ours ResidBase Model=Gemma-2-2B, Probe Type=Residual2026.010.82.350.773.35
Supervised Pinpoint TuningBase Model=Gemma-2-2B2026.010.74.34-0.042.5
Ours ResidModel=Gemma-2-9B, Probe Type=Residual2026.010.641.510.132.2
Ours SAEModel=Gemma-2-9B, Probe Type=SAE2026.010.421.80.543.43
Supervised Pinpoint TuningModel=Gemma-2-9B2026.010.333.640.672.3
Ours SAEBase Model=Gemma-2-2B, Probe Type=SAE2026.010.232.680.663.3
Untrained Gemma-2-2BBase Model=Gemma-2-2B2026.01-0.284.390.532.9
Synthetic Data InterventionModel=Gemma-2-9B2026.01-0.895.220.993.55
Synthetic Data InterventionBase Model=Gemma-2-2B2026.01-1.82-0.36-0.743.52