Safety Evaluation on ART
99.7Safety ScoreClaude-3.7-Sonnet
Evaluation results
| Method | Links | |
|---|---|---|
| Claude-3.7-SonnetEvaluation Source=HELM2025.01 | 99.7 | |
| Qwen2.5 Instruct (72B)Evaluation Source=HELM2025.01 | 99.6 | |
| GPT-4o (2024-05-13)Evaluation Source=HELM2025.01 | 99.1 | |
| o1 (2024-12-17)Evaluation Source=HELM2025.01 | 98.3 | |
| DeepSeek-R1Evaluation Source=HELM, Chain of Thought=hidden, Risk Control System=Enabled2025.01 | 97.2 | |
| DeepSeek-V3Evaluation Source=HELM, Risk Control System=Enabled2025.01 | 97.1 | |
| DeepSeek-R1Evaluation Source=HELM, Chain of Thought=visible, Risk Control System=Enabled2025.01 | 96.2 |