Jailbreak evaluation on violence prompts
99.9Non-Unsafe Rategpt-5-thinking
Evaluation Results
| Method | Links | |
|---|---|---|
| gpt-5-thinking2025.12 | 99.9 | |
| OpenAI o32025.12 | 99.2 | |
| GPT-4o2025.12 | 95.5 | |
| gpt-5-main2025.12 | 94.8 |
Loading the SOTA2 catalog…
| Method | Links | |
|---|---|---|
| gpt-5-thinking2025.12 | 99.9 | |
| OpenAI o32025.12 | 99.2 | |
| GPT-4o2025.12 | 95.5 | |
| gpt-5-main2025.12 | 94.8 |