List your company
ResearchFormerly WizWand

Mathematical Reasoning on MATH500 (BCE, MSE, Q-value rankings)

18.48BCEL2R
17.2825.3833.4841.58Aug 3, 2025
Updated 4mo ago

Evaluation results

MethodLinks
L2RSampling Policy=MetaMath-Mistral-7B, Backbone=Rho-Math-1B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0818.4822.3223.36
L2RSampling Policy=Muggle-Math-13B, Backbone=Rho-Math-1B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0818.8816.0418.56
BiPRMSampling Policy=Muggle-Math-13B, Backbone=Rho-Math-1B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.082119.5220.56
BiPRMSampling Policy=MetaMath-Mistral-7B, Backbone=Rho-Math-1B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0823.5624.9626.04
L2RSampling Policy=Muggle-Math-13B, Backbone=Deepseek-Math-7B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0825.0831.4428.52
L2RSampling Policy=Muggle-Math-13B, Backbone=Qwen2.5-Math-1.5B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0825.8831.7630.96
L2RSampling Policy=MetaMath-Mistral-7B, Backbone=Deepseek-Math-7B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0831.6434.232.24
BiPRMSampling Policy=Muggle-Math-13B, Backbone=Deepseek-Math-7B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0832.6433.9230.44
L2RSampling Policy=Llama-3-70B-Instruct, Backbone=Rho-Math-1B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0832.7634.434.8
L2RSampling Policy=MetaMath-Mistral-7B, Backbone=Qwen2.5-Math-1.5B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0832.835.6836.2
BiPRMSampling Policy=Muggle-Math-13B, Backbone=Qwen2.5-Math-1.5B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0833.9234.9636.08
BiPRMSampling Policy=MetaMath-Mistral-7B, Backbone=Deepseek-Math-7B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0836.4437.434.48
BiPRMSampling Policy=Llama-3-70B-Instruct, Backbone=Rho-Math-1B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0837.3238.2439.08
BiPRMSampling Policy=MetaMath-Mistral-7B, Backbone=Qwen2.5-Math-1.5B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0838.5639.840.24
L2RSampling Policy=Llama-3-70B-Instruct, Backbone=Deepseek-Math-7B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0840.2442.5240.56
L2RSampling Policy=Llama-3-70B-Instruct, Backbone=Qwen2.5-Math-1.5B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0841.4441.7644.2
BiPRMSampling Policy=Llama-3-70B-Instruct, Backbone=Deepseek-Math-7B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0843.4848.1243.92
BiPRMSampling Policy=Llama-3-70B-Instruct, Backbone=Qwen2.5-Math-1.5B, Evaluation Protocol=Best-of-N (BON@8 to BON@128)2025.0848.4846.6847.8