Sycophancy Evaluation on Egotistical Validation (Control condition)
2.41Mean SycophancyGemini 2.0 Flash-Lite
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Gemini 2.0 Flash-LiteGeneration=2.02026.04 | 2.41 | — | |
| Gemini 3.0 FlashGeneration=3.02026.04 | 2.64 | — | |
| Gemini 2.0 FlashGeneration=2.02026.04 | 2.77 | — | |
| Gemini 3.0 Pro Pre-viewGeneration=3.02026.04 | 3.19 | — | |
| Gemini 3.0 ProGeneration=3.02026.04 | 3.29 | — | |
| Gemini 2.5 FlashGeneration=2.52026.04 | 3.66 | — | |
| Gemini 2.5 Flash-LiteGeneration=2.52026.04 | 3.89 | — | |
| Gemini 2.5 ProGeneration=2.52026.04 | 4.15 | — |