OOD safety category inference (Stage 2) on ProGuard Text
32.94Reward-meanProGuard-3B
Evaluation Results
| Method | Links | |
|---|---|---|
| ProGuard-3B2025.12 | 32.94 | |
| ProGuard-7B2025.12 | 32.59 | |
| Gemini2.5-Flash2025.12 | 11.26 | |
| GPT4o-mini2025.12 | 0.12 |
Loading the SOTA2 catalog…
| Method | Links | |
|---|---|---|
| ProGuard-3B2025.12 | 32.94 | |
| ProGuard-7B2025.12 | 32.59 | |
| Gemini2.5-Flash2025.12 | 11.26 | |
| GPT4o-mini2025.12 | 0.12 |