Content Safety on COLD (test)
74.8Macro F1-ScoreGPT-3.5
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-3.52026.06 | 74.8 | |
| Qwen2-72BParameters=72B2026.06 | 73.5 | |
| Yuvion-32BParameters=32B2026.06 | 72.6 | |
| Qwen1.5-72BParameters=72B2026.06 | 71 | |
| DeepSeek-67BParameters=67B2026.06 | 70.8 | |
| Yuvion-8BParameters=8B2026.06 | 69.6 | |
| Qwen2-32BParameters=32B2026.06 | 64.4 | |
| Qwen2-8BParameters=8B2026.06 | 61.2 |