Adversarial Attack on HEx-PHI out
67.29Attack Success Rate (OM)Mistral-7B-Instruct-v0.3
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Mistral-7B-Instruct-v0.3Attack=Prefilling2026.06 | 67.29 | 59.48 | |
| Mistral-7B-Instruct-v0.3Attack=I-GCG2026.06 | 19.93 | 15.61 | |
| Mistral-7B-Instruct-v0.3Attack=PAIR2026.06 | 10.82 | 10.04 | |
| Mistral-7B-Instruct-v0.3 + Trajectory AlignmentAttack=Prefilling2026.06 | 5.42 | 4.09 | |
| Mistral-7B-Instruct-v0.3 + Trajectory AlignmentAttack=I-GCG2026.06 | 2.97 | 0 | |
| Mistral-7B-Instruct-v0.3 + Trajectory AlignmentAttack=PAIR2026.06 | 0 | 0.37 |