List your company
ResearchFormerly WizWand

Mathematical Reasoning on AIME '25 (Mean Accuracy, Standard Deviation)

84Mean AccuracyHermes@1
16.08833.71951.3568.981Nov 24, 2025
Updated 5mo ago

Evaluation results

MethodLinks
Hermes@1Model=OpenAI o3-mini2025.118484
CoT@5+Safe*Model=OpenAI o3-mini2025.1180.780.7
CoT@5+SafeModel=OpenAI o3-mini2025.118080
Hermes@1Model=DeepSeek-V3.22025.1174.774.7
CoT@5+SkyworkModel=OpenAI o3-mini2025.117272
CoT@5+ArmoRMModel=OpenAI o3-mini2025.117272
CoT@1Model=OpenAI o3-mini2025.1171.371.3
CoT@5+ShepherdModel=OpenAI o3-mini2025.117070
Hermes@1Model=DeepSeek-V3.12025.1166.766.7
CoT@5+RLHFlowModel=OpenAI o3-mini2025.1165.365.3
CoT@5+MajorityModel=OpenAI o3-mini2025.1164.764.7
CoT@5+SkyworkModel=DeepSeek-V3.22025.116060
CoT@5+MajorityModel=DeepSeek-V3.22025.1158.758.7
CoT@5+ShepherdModel=DeepSeek-V3.12025.1157.357.3
CoT@5+ShepherdModel=DeepSeek-V3.22025.1157.357.3
CoT@5+ArmoRMModel=DeepSeek-V3.12025.1155.355.3
CoT@5+ArmoRMModel=DeepSeek-V3.22025.1155.355.3
CoT@5+RLHFlowModel=DeepSeek-V3.22025.1153.353.3
CoT@5+Safe*Model=DeepSeek-V3.22025.1152.752.7
CoT@5+SkyworkModel=DeepSeek-V3.12025.115252
CoT@5+RLHFlowModel=DeepSeek-V3.12025.115050
CoT@5+Safe*Model=DeepSeek-V3.12025.115050
CoT@1Model=DeepSeek-V3.22025.1149.349.3
CoT@1Model=DeepSeek-V3.12025.114848
CoT@5+MajorityModel=DeepSeek-V3.12025.114848
CoT@5+SafeModel=DeepSeek-V3.22025.1146.746.7
CoT@5+SafeModel=DeepSeek-V3.12025.1145.345.3
CoT@5+SkyworkModel=Qwen3-8B2025.1130.730.7
CoT@5+ArmoRMModel=Qwen3-8B2025.1130.730.7
Hermes@1Model=Qwen3-8B2025.112828
CoT@5+ShepherdModel=Qwen3-8B2025.1125.325.3
CoT@5+RLHFlowModel=Qwen3-8B2025.112424
CoT@5+SafeModel=Qwen3-8B2025.112222
CoT@5+Safe*Model=Qwen3-8B2025.1121.321.3
CoT@5+MajorityModel=Qwen3-8B2025.112020
CoT@1Model=Qwen3-8B2025.1118.718.7