List your company

Mathematics on GSM8K (test)

84.99AccuracyQwen3-4B-Base
62.224468.134774.04579.9553Jun 6, 2026

Evaluation results

MethodLinks
Qwen3-4B-BaseNumber of shots=42026.0684.99
FReDA-4BNumber of shots=4, Evaluation protocol=Best-of-N2026.0684.15
FReDA-4BNumber of shots=4, Evaluation protocol=Self-refine2026.0683.55
Dream-7B-BaseNumber of shots=42026.0681
TiDAR-8BNumber of shots=4, Evaluation protocol=Trust Diff2026.0680.44
TiDAR-8BNumber of shots=4, Evaluation protocol=Trust AR2026.0679.83
Qwen2.5-3B-BaseNumber of shots=42026.0679.1
BlockDiff-4BNumber of shots=42026.0677.33
LLaDA-8B-BaseNumber of shots=42026.0674.37
LLaDA-MoE-7B-A1B-BaseNumber of shots=42026.0666.41
DiffuLLaMA-7BNumber of shots=4, Note=fine-tuned setting2026.0663.1