Share your thoughts, 1 month free Claude Pro on usSee more

STEM Question Answering on MQA

45First-Token Accuracy

Phi-4-14B

Updated 13d ago

Evaluation Results

Method	Links
Phi-4-14B 2025.05		45
Qwen-2-7B 2025.05		36.5
Gemma-2-9B 2025.05		34
Llama-3.1-8B 2025.05		33.8
Qwen-2-7B 2025.05		29.8
Qwen-2-7B 2025.05		29.8
Gemma-2-9B 2025.05		27.8
Gemma-2-9B 2025.05		27.8
Mistral-Nemo-12B 2025.05		25.4
Gemma-7B 2025.05		25.1
Phi-4-14B 2025.05		25
Phi-4-14B 2025.05		24.9
Ministral-8B 2025.05		24.5
Zephyr-7B 2025.05		24.3
Gemma-7B 2025.05		24.1
Llama-3.1-8B 2025.05		24
Llama-3.1-8B 2025.05		24
Gemma-7B 2025.05		24
Mistral-Nemo-12B 2025.05		23.5
Mistral-Nemo-12B 2025.05		23.5
Zephyr-7B 2025.05		23.3
Zephyr-7B 2025.05		23.3
Ministral-8B 2025.05		23.1
Ministral-8B 2025.05		23.1