Multiple-choice Question Answering on KNIGHT Biology L3
91.6AccuracyHuman
Evaluation results
| Method | Links | |
|---|---|---|
| Humann=2002026.02 | 91.6 | |
| GPT-4o2026.02 | 87.11 | |
| Mistral Large2026.02 | 86.84 | |
| Llama3-70B-InstructSize=70B, Type=Instruct2026.02 | 86.08 | |
| Claude 3 Haiku2026.02 | 83.4 | |
| Qwen1.5 (1.8B)Size=1.8B2026.02 | 72.89 | |
| Gemma (2B)Size=2B2026.02 | 31.27 |