Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Sycophancy Evaluation on TruthfulQA (adversarial)
Loading...
1
Sycophantic Response Count
Silicon Mirror
0.8
2.15
3.5
4.85
Apr 1, 2026
Sycophantic Response Count
Sycophancy Rate
Sycophancy Rate 95% CI (Lower Bound)
Relative Reduction vs Vanilla
Updated 17d ago
Evaluation Results
Method
Method
Links
Sycophantic Response Count
Sycophancy Rate
Sycophancy Rate 95% CI (Lower Bound)
Relative Reduction vs Vanilla
Silicon Mirror
Condition=full pipelin...
2026.04
1
2
0.1
83.3
Static guardrails
Condition="be truthful...
2026.04
2
4
0.5
66.7
Vanilla
Condition=no intervent...
2026.04
6
12
4.5
-
Feedback
Search any
task
Search any
task