List your company
ResearchFormerly WizWand

Reasoning on LQA

40.6First-Token AccuracyZephyr-7B
25.5229.43533.3537.265May 21, 2025
Updated 7mo ago

Evaluation results

MethodLinks
Zephyr-7BStrategy=Prefilling2025.0540.6
Gemma-2-9BStrategy=Prefilling2025.0535.8
Qwen-2-7BStrategy=Prefilling2025.0535.5
Ministral-8BStrategy=Prefilling2025.0534.6
Phi-4-14BStrategy=Prefilling2025.0534.6
Phi-4-14BStrategy=Prompt Engineering2025.0534.1
Llama-3.1-8BStrategy=Prefilling2025.0534
Qwen-2-7BStrategy=FTP2025.0533
Gemma-2-9BStrategy=Prompt Engineering2025.0532.1
Zephyr-7BStrategy=FTP2025.0532.1
Mistral-Nemo-12BStrategy=Prefilling2025.0532
Mistral-Nemo-12BStrategy=Prompt Engineering2025.0531.9
Phi-4-14BStrategy=FTP2025.0531.9
Qwen-2-7BStrategy=Prompt Engineering2025.0531.6
Gemma-2-9BStrategy=FTP2025.0531
Zephyr-7BStrategy=Prompt Engineering2025.0530.1
Ministral-8BStrategy=Prompt Engineering2025.0529
Llama-3.1-8BStrategy=FTP2025.0528.9
Mistral-Nemo-12BStrategy=FTP2025.0528.7
Llama-3.1-8BStrategy=Prompt Engineering2025.0528.4
Ministral-8BStrategy=FTP2025.0528.4
Gemma-7BStrategy=Prompt Engineering2025.0527.3
Gemma-7BStrategy=FTP2025.0526.4
Gemma-7BStrategy=Prefilling2025.0526.1