Stereotyping Evaluation on DecodingTrust Stereotype (DTS)
10.7Stereotype FractionGemini-Pro
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Gemini-ProModel=Gem-Pro2024.07 | 10.7 | 37.2 | 21.7 | — | |
| GPT-4oModel=GPT-4o2024.07 | 11.8 | 40.1 | 23.7 | — | |
| GPT-4o-miniModel=GPT-4o-m2024.07 | 23 | 46.6 | 30.1 | — | |
| Gemini-Flash-LiteModel=Gem-Fl-Lt2024.07 | 24.6 | 50.1 | 24.1 | — | |
| Gemini-FlashModel=Gem-Fl2024.07 | 28.4 | 38.2 | 24.5 | — | |
| Gem-Fl2024.07 | — | — | — | 0.5 | |
| Gem-Fl-Lt2024.07 | — | — | — | 1.2 | |
| Gem-Pro2024.07 | — | — | — | 0.5 | |
| GPT-4o2024.07 | — | — | — | 0.4 | |
| GPT-4o-m2024.07 | — | — | — | 1.3 |