SOTA Mathematical Reasoning on Math ID GSM8k ProofNet and PapersWithCode

99.2GSM8k Accuracy

Qwen3-8B pass@N (Upper Bound)

Updated 1mo ago

Evaluation Results

Method
Qwen3-8B pass@N (Upper Bound) 2025.11	99.2	99.3	97.8	-
Qwen2.5-Math-PRM-7B 2025.11	97.8	93.7	76	-
ReProbe, Attn+Logit, DeepSeek-anno 2025.11	97.8	92.7	76.5	-
Qwen3-14B pass@1 2025.11	97.6	93.4	76	-
Majority Voting 2025.11	97.6	-	-	-
Skywork-PRM-1.5B 2025.11	97.6	94.4	76.5	-
Universal-PRM-Qwen2.5-Math-7B 2025.11	97.5	95.7	76	-
ReProbe, Attn+Logit, Self-anno 2025.11	97.5	94.4	73.6	-
Qwen2.5-Math-7B-PRM800k 2025.11	97.3	92.7	74.4	-
RLHFlow-PRM-Deepseek-Data 2025.11	96.4	92.7	71.7	-
RLHFlow-PRM-Mistral-Data 2025.11	96.3	93.7	71.7	-
ReProbe, Hidden States, Self-anno 2025.11	96.1	94.4	78.5	-
ReProbe, Hidden States, DeepSeek-anno 2025.11	96.1	92.7	74.1	-
Qwen3-8B pass@1 (Lower Bound) 2025.11	95.6	92.4	74.1	-
Math-Shepherd-PRM-7B 2025.11	95.5	93	72.8	-
H4-Qwen2.5-PRM-1.5B-0.2 2025.11	95.1	91.7	71.4	-
Qwen3-1.7B pass@N (Upper Bound) 2025.11	93.3	83.1	77.4	84.6
Qwen2.5-Math-PRM-7B 2025.11	85.4	62.6	52.9	67
Universal-PRM-Qwen2.5-Math-7B 2025.11	85	67.3	57	69.8
ReProbe, Hidden States, GPT-OSS-anno 2025.11	85	62.2	57.9	68.4
Math-Shepherd-PRM-7B 2025.11	84.6	67.3	56.2	69.4
MaxProb 2025.11	84	55.6	52.9	64.2
Perplexity 2025.11	84	55.6	52.9	64.2
MaxEntropy 2025.11	83.6	62.2	47.1	64.3
RLHFlow-PRM-Deepseek-Data 2025.11	82.8	64.9	49.6	65.8
RLHFlow-PRM-Mistral-Data 2025.11	82.4	62.6	48.8	64.6
H4-Qwen2.5-PRM-1.5B-0.2 2025.11	78.3	63.7	52.9	65
Qwen3-1.7B pass@1 (Lower Bound) 2025.11	70.1	54.6	47.7	57.5