List your company
ResearchFormerly WizWand

Mathematical Problem Solving on MATH (Latency and Performance)

3.94Average TimeCoT
0.581623.250845.9268.5892Jul 20, 2025
Updated 5mo ago

Evaluation results

MethodLinks
CoTModel=Qwen3-1.7B2025.073.949.91—
LLM As a CriticModel=Qwen3-1.7B, Variant=MUR2025.074.4917.752.29
Guided SearchModel=Qwen3-1.7B, Variant=MUR2025.074.9616.8-35.81
LLM As a CriticModel=Qwen3-1.7B, Variant=SMART2025.075.3826.6611.2
LLM As a CriticModel=Qwen3-1.7B, Variant=Avg uncertainty2025.077.3216.941.48
ϕ-DecodingModel=Qwen3-1.7B, Variant=MUR2025.078.1221.81-21.57
LLM As a CriticModel=Qwen3-1.7B, Variant=Per-Step Scale2025.078.3515.46—
Guided SearchModel=Qwen3-1.7B, Variant=Avg uncertainty2025.078.6822.41-30.2
ϕ-DecodingModel=Qwen3-1.7B, Variant=Avg uncertainty2025.079.2423.55-19.83
CoTModel=Qwen3-4B2025.079.4615.63—
Guided SearchModel=Qwen3-1.7B, Variant=SMART2025.0710.5737.37-15.24
Guided SearchModel=Qwen3-4B, Variant=MUR2025.0710.6934.11-60.24
ϕ-DecodingModel=Qwen3-1.7B, Variant=SMART2025.0711.2634.45-8.93
Guided SearchModel=Qwen3-1.7B, Variant=Per-Step Scale2025.0712.852.61—
LLM As a CriticModel=Qwen3-4B, Variant=MUR2025.0712.8438.41-20.81
ϕ-DecodingModel=Qwen3-1.7B, Variant=Per-Step Scale2025.0715.5243.38—
LLM As a CriticModel=Qwen3-4B, Variant=Avg uncertainty2025.0715.6547.86-11.36
LLM As a CriticModel=Qwen3-4B, Variant=SMART2025.0715.9437.26-21.96
LLM As a CriticModel=Qwen3-4B, Variant=Per-Step Scale2025.0718.5859.22—
Guided SearchModel=Qwen3-4B, Variant=SMART2025.0719.3496.382.03
ϕ-DecodingModel=Qwen3-4B, Variant=Avg uncertainty2025.0719.759.94-35.07
ϕ-DecodingModel=Qwen3-4B, Variant=MUR2025.0720.8358.06-36.95
Guided SearchModel=Qwen3-4B, Variant=Avg uncertainty2025.0721.8362.78-31.57
CoTModel=Qwen3-8B2025.0727.1845.38—
ϕ-DecodingModel=Qwen3-4B, Variant=SMART2025.0728.7591.93-3.08
LLM As a CriticModel=Qwen3-8B, Variant=Avg uncertainty2025.0730.6178.8-0.07
Guided SearchModel=Qwen3-4B, Variant=Per-Step Scale2025.0731.1294.35—
LLM As a CriticModel=Qwen3-8B, Variant=MUR2025.0731.3872.22-6.65
LLM As a CriticModel=Qwen3-8B, Variant=SMART2025.0735.1277.09-1.78
LLM As a CriticModel=Qwen3-8B, Variant=Per-Step Scale2025.0736.4178.87—
ϕ-DecodingModel=Qwen3-4B, Variant=Per-Step Scale2025.0742.5995.01—
Guided SearchModel=Qwen3-8B, Variant=Avg uncertainty2025.0743.28114.81-92.39
Guided SearchModel=Qwen3-8B, Variant=MUR2025.0746.68113.52-93.68
ϕ-DecodingModel=Qwen3-8B, Variant=MUR2025.0751.32160.16-79.94
Guided SearchModel=Qwen3-8B, Variant=SMART2025.0755.64137.16-70.04
ϕ-DecodingModel=Qwen3-8B, Variant=Avg uncertainty2025.0758.35163.97-76.13
Guided SearchModel=Qwen3-8B, Variant=Per-Step Scale2025.0759.32207.2—
ϕ-DecodingModel=Qwen3-8B, Variant=SMART2025.0780.2207.32-32.78
ϕ-DecodingModel=Qwen3-8B, Variant=Per-Step Scale2025.0787.9240.1—