Perturbed Policy Refusal on Health Insurance Portability and Accountability Act (HIPAA)
5,960Refusal RateDEEPSEEK-R1
Evaluation Results
| Method | Links | |
|---|---|---|
| DEEPSEEK-R1Perturbation=Deontic Norm, DEF Stage=DEF 32026.01 | 5,960 | |
| DEEPSEEK-R1Perturbation=Deontic Norm, DEF Stage=DEF 22026.01 | 5,150 | |
| DEEPSEEK-R1Perturbation=Authorization, DEF Stage=DEF 32026.01 | 4,780 | |
| DEEPSEEK-R1Perturbation=Deontic Norm, DEF Stage=DEF 12026.01 | 4,490 | |
| DEEPSEEK-R1Perturbation=Authorization, DEF Stage=DEF 22026.01 | 3,820 | |
| DEEPSEEK-R1Perturbation=Authorization, DEF Stage=DEF 12026.01 | 2,280 | |
| QWEN3-30B-THINKPerturbation=Authorization, DEF Stage=DEF 32026.01 | 1,840 | |
| QWEN3-30B-THINKPerturbation=Deontic Norm, DEF Stage=DEF 12026.01 | 1,400 | |
| QWEN3-30B-THINKPerturbation=Deontic Norm, DEF Stage=DEF 22026.01 | 1,320 | |
| QWEN3-30B-THINKPerturbation=Authorization, DEF Stage=DEF 12026.01 | 1,250 | |
| QWEN3-30B-THINKPerturbation=Deontic Norm, DEF Stage=DEF 32026.01 | 1,100 | |
| QWEN3-30B-THINKPerturbation=Authorization, DEF Stage=DEF 22026.01 | 1,100 | |
| GPT-5-MINIPerturbation=Deontic Norm, DEF Stage=DEF 32026.01 | 590 | |
| GPT-5-MINIPerturbation=Authorization, DEF Stage=DEF 32026.01 | 590 | |
| GPT-5-MINIPerturbation=Authorization, DEF Stage=DEF 22026.01 | 440 | |
| QWEN3-30B-THINKPerturbation=Deontic Norm, DEF Stage=No DEF2026.01 | 290 | |
| DEEPSEEK-R1Perturbation=Deontic Norm, DEF Stage=Correct2026.01 | 240 | |
| DEEPSEEK-R1Perturbation=Authorization, DEF Stage=Correct2026.01 | 240 | |
| GPT-5-MINIPerturbation=Authorization, DEF Stage=DEF 12026.01 | 220 | |
| QWEN3-30B-THINKPerturbation=Authorization, DEF Stage=No DEF2026.01 | 70 | |
| GPT-5-MINIPerturbation=Deontic Norm, DEF Stage=DEF 12026.01 | 70 | |
| GPT-5-MINIPerturbation=Deontic Norm, DEF Stage=DEF 22026.01 | 70 | |
| QWEN3-30B-THINKPerturbation=Deontic Norm, DEF Stage=Correct2026.01 | 0 | |
| QWEN3-30B-THINKPerturbation=Authorization, DEF Stage=Correct2026.01 | 0 | |
| DEEPSEEK-R1Perturbation=Deontic Norm, DEF Stage=No DEF2026.01 | 0 | |
| DEEPSEEK-R1Perturbation=Authorization, DEF Stage=No DEF2026.01 | 0 | |
| GPT-5-MINIPerturbation=Deontic Norm, DEF Stage=Correct2026.01 | 0 | |
| GPT-5-MINIPerturbation=Deontic Norm, DEF Stage=No DEF2026.01 | 0 | |
| GPT-5-MINIPerturbation=Authorization, DEF Stage=Correct2026.01 | 0 | |
| GPT-5-MINIPerturbation=Authorization, DEF Stage=No DEF2026.01 | 0 |