List your company

Mathematical Reasoning on GSM1K

76.1AccuracyT-All
24.72438.06251.464.738May 10, 2026

Evaluation results

MethodLinks
T-AllBackbone=LLaMA-3.1-8B-Instruct2026.0576.1
T-MLPBackbone=LLaMA-3.1-8B-Instruct2026.0575.4
BaseBackbone=LLaMA-3.1-8B-Instruct2026.0573.9
Vanilla SFTBackbone=LLaMA-3.1-8B-Instruct2026.0564.2
T-MLPBackbone=LLaMA-3.2-3B-Instruct2026.0557.5
T-AllBackbone=LLaMA-3.2-3B-Instruct2026.0557.2
BaseBackbone=LLaMA-3.2-3B-Instruct2026.0556.3
T-AllBackbone=Mistral-7B-Instruct2026.0541.4
T-MLPBackbone=Mistral-7B-Instruct2026.0540.8
BaseBackbone=Mistral-7B-Instruct2026.0539.9
Vanilla SFTBackbone=LLaMA-3.2-3B-Instruct2026.0539
Vanilla SFTBackbone=Mistral-7B-Instruct2026.0526.7