List your company

Mathematical Reasoning on GSM8K (Accuracy, Time)

48.3AccuracyFor-Value
41.22843.06444.946.736Aug 13, 2025

Evaluation results

MethodLinks
For-ValueBackbone=Llama-3.1-8B, Selection Ratio=5%2025.0848.30.3
Full (100%)Backbone=Llama-3.1-8B, Selection Ratio=100%2025.0847.8—
For-ValueBackbone=Llama-3.1-8B, Selection Ratio=1%2025.0845.20.3
HyperINFBackbone=Llama-3.1-8B, Selection Ratio=5%2025.0842.82.4
DataInfBackbone=Llama-3.1-8B, Selection Ratio=5%2025.08421.9
HyperINFBackbone=Llama-3.1-8B, Selection Ratio=1%2025.0841.92.4
Hessian-freeBackbone=Llama-3.1-8B, Selection Ratio=5%2025.0841.81.4
DataInfBackbone=Llama-3.1-8B, Selection Ratio=1%2025.0841.71.9
Hessian-freeBackbone=Llama-3.1-8B, Selection Ratio=1%2025.0841.51.4