Multiple-choice Question Answering on Open Quiz Commons Cybersecurity 2024
94.83AccuracyReasoning-trace comparison
Evaluation Results
| Method | Links | |
|---|---|---|
| Reasoning-trace comparisonStrategy=Reasoning-combined, Judge Model=Llama 3.22026.06 | 94.83 | |
| Reasoning-trace comparisonStrategy=Reasoning-combined, Judge Model=Phi-32026.06 | 93.1 | |
| Reasoning-trace comparisonStrategy=Reasoning-combined, Judge Model=Gemma-22026.06 | 93.1 | |
| Gemma-2Strategy=Original Prediction2026.06 | 87.93 | |
| Phi-3Strategy=Original Prediction2026.06 | 60.34 |