Robustness against priming vulnerability on JBB-Behaviors (test)
0ASR (Guardrail Model)
RA (ours)
Updated 4d ago
Evaluation Results
| Method | Links | ||
|---|---|---|---|
2025.10 | 0 | 130 | |
2025.10 | 0 | 0 | |
2025.10 | 0 | 170 | |
2025.10 | 0 | 430 | |
2025.10 | 70 | 100 | |
2025.10 | 70 | 500 | |
2025.10 | 100 | 230 | |
2025.10 | 200 | 670 | |
2025.10 | 400 | 570 | |
2025.10 | 530 | 1,230 | |
2025.10 | 1,170 | 1,930 | |
2025.10 | 2,730 | 1,770 | |
2025.10 | 2,830 | 1,430 | |
2025.10 | 2,830 | 2,330 | |
2025.10 | 2,870 | 2,600 | |
2025.10 | 4,300 | 4,830 | |
2025.10 | 5,800 | 3,870 | |
2025.10 | 7,400 | 6,100 | |
2025.10 | 7,630 | 3,830 | |
2025.10 | 8,730 | 6,070 |