STEM Question Answering on MQA
45First-Token AccuracyPhi-4-14B
Evaluation Results
| Method | Links | |
|---|---|---|
| Phi-4-14BStrategy=Prefilling2025.05 | 45 | |
| Qwen-2-7BStrategy=Prefilling2025.05 | 36.5 | |
| Gemma-2-9BStrategy=Prefilling2025.05 | 34 | |
| Llama-3.1-8BStrategy=Prefilling2025.05 | 33.8 | |
| Qwen-2-7BStrategy=FTP2025.05 | 29.8 | |
| Qwen-2-7BStrategy=Prompt Engineering2025.05 | 29.8 | |
| Gemma-2-9BStrategy=FTP2025.05 | 27.8 | |
| Gemma-2-9BStrategy=Prompt Engineering2025.05 | 27.8 | |
| Mistral-Nemo-12BStrategy=Prefilling2025.05 | 25.4 | |
| Gemma-7BStrategy=Prefilling2025.05 | 25.1 | |
| Phi-4-14BStrategy=FTP2025.05 | 25 | |
| Phi-4-14BStrategy=Prompt Engineering2025.05 | 24.9 | |
| Ministral-8BStrategy=Prefilling2025.05 | 24.5 | |
| Zephyr-7BStrategy=Prefilling2025.05 | 24.3 | |
| Gemma-7BStrategy=FTP2025.05 | 24.1 | |
| Llama-3.1-8BStrategy=FTP2025.05 | 24 | |
| Llama-3.1-8BStrategy=Prompt Engineering2025.05 | 24 | |
| Gemma-7BStrategy=Prompt Engineering2025.05 | 24 | |
| Mistral-Nemo-12BStrategy=FTP2025.05 | 23.5 | |
| Mistral-Nemo-12BStrategy=Prompt Engineering2025.05 | 23.5 | |
| Zephyr-7BStrategy=FTP2025.05 | 23.3 | |
| Zephyr-7BStrategy=Prompt Engineering2025.05 | 23.3 | |
| Ministral-8BStrategy=FTP2025.05 | 23.1 | |
| Ministral-8BStrategy=Prompt Engineering2025.05 | 23.1 |