List your company

STEM Question Answering on MQA

45First-Token AccuracyPhi-4-14B
22.22428.13734.0539.963May 21, 2025

Evaluation results

MethodLinks
Phi-4-14BStrategy=Prefilling2025.0545
Qwen-2-7BStrategy=Prefilling2025.0536.5
Gemma-2-9BStrategy=Prefilling2025.0534
Llama-3.1-8BStrategy=Prefilling2025.0533.8
Qwen-2-7BStrategy=FTP2025.0529.8
Qwen-2-7BStrategy=Prompt Engineering2025.0529.8
Gemma-2-9BStrategy=FTP2025.0527.8
Gemma-2-9BStrategy=Prompt Engineering2025.0527.8
Mistral-Nemo-12BStrategy=Prefilling2025.0525.4
Gemma-7BStrategy=Prefilling2025.0525.1
Phi-4-14BStrategy=FTP2025.0525
Phi-4-14BStrategy=Prompt Engineering2025.0524.9
Ministral-8BStrategy=Prefilling2025.0524.5
Zephyr-7BStrategy=Prefilling2025.0524.3
Gemma-7BStrategy=FTP2025.0524.1
Llama-3.1-8BStrategy=FTP2025.0524
Llama-3.1-8BStrategy=Prompt Engineering2025.0524
Gemma-7BStrategy=Prompt Engineering2025.0524
Mistral-Nemo-12BStrategy=FTP2025.0523.5
Mistral-Nemo-12BStrategy=Prompt Engineering2025.0523.5
Zephyr-7BStrategy=FTP2025.0523.3
Zephyr-7BStrategy=Prompt Engineering2025.0523.3
Ministral-8BStrategy=FTP2025.0523.1
Ministral-8BStrategy=Prompt Engineering2025.0523.1