Jailbreak evaluation on illicit non-violent crime prompts
99.5Not Unsafe Rategpt-5-thinking
Evaluation Results
| Method | Links | |
|---|---|---|
| gpt-5-thinking2025.12 | 99.5 | |
| OpenAI o32025.12 | 98.5 | |
| GPT-4o2025.12 | 93.7 | |
| gpt-5-main2025.12 | 93.4 |
Loading the SOTA2 catalog…
| Method | Links | |
|---|---|---|
| gpt-5-thinking2025.12 | 99.5 | |
| OpenAI o32025.12 | 98.5 | |
| GPT-4o2025.12 | 93.7 | |
| gpt-5-main2025.12 | 93.4 |