List your company

Mathematical Reasoning on MATH v1 (test)

83.8AccuracyPass@k
20.56836.98453.469.816Jul 15, 2024Nov 6, 2024Feb 28, 2025Jun 23, 2025Oct 15, 2025Feb 6, 2026Jun 1, 2026

Evaluation results

MethodLinks
Pass@kModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=322026.0683.8
Pass@kModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=162026.0678.6
Large greedyModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=82026.0678
Large greedyModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=162026.0678
Large greedyModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=322026.0678
Pass@kModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=82026.0671
PRM guided searchModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=322026.0669.8
CGSModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=32, Chunk Length (L)=202026.0668.8
Large greedyModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=82026.0668
Large greedyModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=162026.0668
Large greedyModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=322026.0668
LGSModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=32, Chunk Length (L)=202026.0667.4
PRM guided searchModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=162026.0666.4
Pass@kModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=322026.0666.4
CGSModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=16, Chunk Length (L)=202026.0664.6
PRM guided searchModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=82026.0664.2
Borda count (large)Model Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=322026.0663.4
CGSModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=8, Chunk Length (L)=202026.0663.2
Borda count (small)Model Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=322026.0663.2
Majority@kModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=322026.0663
LGSModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=16, Chunk Length (L)=202026.0661.8
Pass@kModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=162026.0660.8
LGSModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=8, Chunk Length (L)=202026.0660
Borda count (large)Model Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=162026.0660
Majority@kModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=162026.0659.6
Borda count (small)Model Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=162026.0659.6
Borda count (large)Model Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=82026.0657.6
Borda count (small)Model Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=82026.0657
Majority@kModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=82026.0656.6
Best-of-NModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=82026.0655.8
Self-Certainty (large)Model Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=322026.0653.6
Best-of-NModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=162026.0653.4
Qwen2-7BInstruction-tuned=true2024.0752.9
Small greedyModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=82026.0652.8
Small greedyModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=162026.0652.8
Small greedyModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=322026.0652.8
Self-Certainty (small)Model Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=82026.0652.6
Self-Certainty (small)Model Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=322026.0652.6
Best-of-NModel Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=322026.0652.6
Self-Certainty (large)Model Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=82026.0651.2
Self-Certainty (small)Model Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=162026.0651
GLM-4-9BInstruction-tuned=true2024.0750.6
Pass@kModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=82026.0649.6
Self-Certainty (large)Model Pair=Qwen2.5-1.5B → Qwen2.5-32B, Sample Size (k)=162026.0649.2
Yi-1.5-9BInstruction-tuned=true2024.0747.7
PRM guided searchModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=322026.0646.8
CGSModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=32, Chunk Length (L)=202026.0646.4
CGSModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=16, Chunk Length (L)=202026.0644.4
LGSModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=32, Chunk Length (L)=202026.0644.2
PRM guided searchModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=162026.0643.4
PRM guided searchModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=82026.0640.2
CGSModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=8, Chunk Length (L)=202026.0639.8
LGSModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=16, Chunk Length (L)=202026.0638.8
Borda count (small)Model Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=322026.0637.8
Borda count (large)Model Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=322026.0637.8
LGSModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=8, Chunk Length (L)=202026.0636.6
Majority@kModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=322026.0636
Borda count (large)Model Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=162026.0634.6
Majority@kModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=162026.0634.2
Borda count (small)Model Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=162026.0633.4
Self-Certainty (large)Model Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=322026.0633.4
Best-of-NModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=322026.0632.2
Borda count (large)Model Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=82026.0631.4
Best-of-NModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=82026.0630.6
Llama-3-8BInstruction-tuned=true2024.0730
Self-Certainty (large)Model Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=162026.0630
Self-Certainty (large)Model Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=82026.0629.8
Best-of-NModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=162026.0629
Borda count (small)Model Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=82026.0628.6
Majority@kModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=82026.0628.2
Self-Certainty (small)Model Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=322026.0625.8
Small greedyModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=82026.0625.2
Small greedyModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=162026.0625.2
Small greedyModel Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=322026.0625.2
Self-Certainty (small)Model Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=162026.0624.4
Qwen1.5-7BInstruction-tuned=true2024.0723.2
Self-Certainty (small)Model Pair=Llama-3.2-1B → Llama-3.1-70B, Sample Size (k)=82026.0623