| bias-detection-on-rt-inod-bias | GPT-4 | #1 | Best-of: 0.5 |
| bias-detection-on-rt-inod-bias | Gemma | #2 | Best-of: 0.41 |
| bias-detection-on-rt-inod-bias | Baseline | #3 | Best-of: 0.41 |
| bias-detection-on-rt-inod-bias | Mistral | #4 | Best-of: 0.36 |
| bias-detection-on-rt-inod-bias | Llama2 | #5 | Best-of: 0.34 |
| dialogue-safety-prediction-on-rt-inod-jailbreaking | Baseline | #1 | Best-of: 0.92 |
| dialogue-safety-prediction-on-rt-inod-jailbreaking | GPT-4 | #2 | Best-of: 0.91 |
| dialogue-safety-prediction-on-rt-inod-jailbreaking | Gemma | #3 | Best-of: 0.91 |
| dialogue-safety-prediction-on-rt-inod-jailbreaking | Mistral | #4 | Best-of: 0.87 |
| dialogue-safety-prediction-on-rt-inod-jailbreaking | Llama2 | #5 | Best-of: 0.86 |