Gender bias evaluation on RealWorldQuestioning Jobs Recommendations 1.0
70.54Male More InformationLlama-3
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Llama-3Judge Model=ChatGPT-4o, Iterations=12025.05 | 70.54 | 25.58 | 3.87 | 0.14 | 0 | |
| DeepSeek-R1Judge Model=ChatGPT-4o, Iterations=12025.05 | 59.68 | 40.31 | 0 | 0.45 | 0.002 | |
| ChatGPT-4-turboJudge Model=ChatGPT-4o, Iterations=12025.05 | 54.26 | 44.18 | 1.55 | 0.66 | 0.13 | |
| ChatGPT-3.5-turboJudge Model=ChatGPT-4o, Iterations=12025.05 | 46.51 | 50.38 | 3.1 | 1.17 | 0.62 |