List your company
ResearchFormerly WizWand

Mathematical Reasoning on GSM+ (Accuracy)

62.29Accuracy (GSM+)Qwen-2.5-3B-instruct
1.980417.637733.29548.9523Jul 2, 2025Aug 25, 2025Oct 19, 2025Dec 13, 2025Feb 5, 2026Apr 1, 2026May 26, 2026
Updated 4mo ago

Evaluation results

MethodLinks
Qwen-2.5-3B-instruct2025.0762.29
BBoxERBase Model=Qwen-2.5-3B-instruct, Budget (b)=150, Optimization Algorithm=D-CMA, Update Type=normalization layers2025.0762.27
BBoxERBase Model=Qwen-2.5-3B-instruct, Budget (b)=300, Optimization Algorithm=D-CMA, Update Type=normalization layers2025.0761.94
BBoxERBase Model=Llama3.1-8B-Instruct, Budget (b)=300, Optimization Algorithm=D-CMA, Update Type=normalization layers2025.0755.33
BBoxERBase Model=Llama3.1-8B-Instruct, Budget (b)=150, Optimization Algorithm=D-CMA, Update Type=normalization layers2025.0755.11
Llama3.1-8B-Instruct2025.0754.37
MobileMoE-LN_act=922M, N_total=5.3B2026.0550.2
Qwen3.5 2BN_act=1.9B, N_total=1.9B2026.0542.4
MobileMoE-MN_act=528M, N_total=2.8B2026.0542.3
BBoxERBase Model=Llama3.1-8B, Budget (b)=300, Optimization Algorithm=D-CMA, Update Type=normalization layers2025.0737.88
BBoxERBase Model=Llama3.1-8B, Budget (b)=150, Optimization Algorithm=D-CMA, Update Type=normalization layers2025.0737.28
Llama3.1-8B2025.0736.5
MobileMoE-SN_act=272M, N_total=1.3B2026.0528.9
OLMoE-1B-7BN_act=1.3B, N_total=6.9B2026.0528.1
Llama 3.2 1BN_act=1.2B, N_total=1.2B2026.0527.8
SmolLM2 1.7BN_act=1.7B, N_total=1.7B2026.0526.6
Qwen3.5 0.8BN_act=749M, N_total=749M2026.0526.5
Gemma 3 1BN_act=1.0B, N_total=1.0B2026.0525.2
OLMo 2 1BN_act=1.5B, N_total=1.5B2026.0525.1
MobileLLM-ProN_act=1.1B, N_total=1.1B2026.0517
SmolLM2 360MN_act=362M, N_total=362M2026.054.6
Gemma 3 270MN_act=270M, N_total=270M2026.054.3