Share your thoughts, 1 month free Claude Pro on usSee more

Best-of-N Reranking on Average of 7 Benchmarks (AIME24, LeetCode) (test)

52Average Accuracy

DeepSeek-R1-Distill-Qwen-32B

Updated 14d ago

Evaluation Results

Method	Links
DeepSeek-R1-Distill-Qwen-32B 2025.03		52
DeepSeek-R1-Distill-Qwen-32B 2025.03		51.1
Qwen2.5-Math-PRM-72B 2025.03		50.6
DeepSeek-R1-Distill-Qwen-32B 2025.03		50.3
DeepSeek-R1-Distill-Qwen-7B 2025.03		50.1
Skywork-o1-Open-PRM-Qwen-2.5-7B 2025.03		49.9
Qwen2.5-Math-PRM-7B 2025.03		48.7
DeepSeek-R1-Distill-Qwen-7B 2025.03		48.7
DeepSeek-R1-Distill-Qwen-32B 2025.03		48.5
DeepSeek-R1-Distill-Qwen-7B 2025.03		48.3
DeepSeek-R1-Distill-Qwen-32B 2025.03		48.1
Skywork-o1-Open-PRM-Qwen-2.5-1.5B 2025.03		46.7
DeepSeek-R1-Distill-Qwen-7B 2025.03		46.6
Skywork-Reward-Gemma-2-27B-v0.2 2025.03		45.6
Qwen2.5-72B-Instruct 2025.03		45.6
DeepSeek-R1-Distill-Qwen-7B 2025.03		45.6
Skywork-Reward-Llama-3.1-8B-v0.2 2025.03		45.5
Skywork-Reward-Gemma-2-27B-v0.2 2025.03		45.5
Skywork-Reward-Gemma-2-27B-v0.2 2025.03		45.4
Skywork-Reward-Llama-3.1-8B-v0.2 2025.03		45.2
Skywork-Reward-Llama-3.1-8B-v0.2 2025.03		44.8
Skywork-Reward-Gemma-2-27B-v0.2 2025.03		44.8
Skywork-Reward-Llama-3.1-8B-v0.2 2025.03		44.6
Qwen2.5-Math-7B-PRM800K 2025.03		44.6
DeepSeek-R1-Distill-Qwen-32B 2025.03		44.4
Skywork-Reward-Llama-3.1-8B-v0.2 2025.03		43.4
Skywork-Reward-Gemma-2-27B-v0.2 2025.03		43.4
DeepSeek-R1-Distill-Qwen-7B 2025.03		43.3
math-shepherd-mistral-7b-prm 2025.03		43.2
Llama3-8B-CLoud-RM 2025.03		42.5
Skywork-Reward-Gemma-2-27B-v0.2 2025.03		41.8
Skywork-Reward-Llama-3.1-8B-v0.2 2025.03		41.4
prometheus-8x7b-v2.0 2025.03		40.8
prometheus-7b-v2.0 2025.03		39.7
Skywork-Reward-Llama-3.1-8B-v0.2 2025.03		38.2
Skywork-Reward-Gemma-2-27B-v0.2 2025.03		38.2
math-shepherd-mistral-7b-prm 2025.03		38.2
DeepSeek-R1-Distill-Qwen-7B 2025.03		38.2
DeepSeek-R1-Distill-Qwen-7B 2025.03		38.2
DeepSeek-R1-Distill-Qwen-32B 2025.03		38.2
RLHFlow/Llama3.1-8B-PRM-Deepseek 2025.03		37.8
RLHFlow/Llama3.1-8B-PRM-Mistral 2025.03		35.5