List your company

Mathematical Reasoning on AIME 25 (Avg@32, Average)

34.38Avg@32BuPO
-1.37527.907417.1926.4726Dec 22, 2025

Evaluation results

MethodLinks
BuPOBackbone=Qwen3-8B2025.1234.3866.36
GRPOBackbone=Qwen3-8B2025.1233.5464.23
RLOOBackbone=Qwen3-8B2025.1233.0263.36
BuPOBackbone=Qwen3-4B2025.1231.1558.51
Reinforce++Backbone=Qwen3-8B2025.1231.1560.41
GRPOBackbone=Qwen3-4B2025.1228.8555.08
PPOBackbone=Qwen3-4B2025.1227.655.22
RLOOBackbone=Qwen3-4B2025.1224.7954
PPOBackbone=Qwen3-8B2025.1222.658.41
VanillaBackbone=Qwen3-8B2025.1219.1748.49
Reinforce++Backbone=Qwen3-4B2025.1218.6545.03
VanillaBackbone=Qwen3-4B2025.1218.647.44
BuPOBackbone=Llama-OctoThinker-8B-Base2025.126.7727.79
Reinforce++Backbone=Llama-OctoThinker-8B-Base2025.123.7526.43
GRPOBackbone=Llama-OctoThinker-8B-Base2025.122.1924.11
RLOOBackbone=Llama-OctoThinker-8B-Base2025.121.5622.18
PPOBackbone=Llama-OctoThinker-8B-Base2025.121.0422.82
BuPOBackbone=Llama-OctoThinker-3B-Base2025.120.4219.59
PPOBackbone=Llama-OctoThinker-3B-Base2025.120.3116.69
RLOOBackbone=Llama-OctoThinker-3B-Base2025.120.2117.84
Reinforce++Backbone=Llama-OctoThinker-3B-Base2025.120.15.27
GRPOBackbone=Llama-OctoThinker-3B-Base2025.120.118.58
VanillaBackbone=Llama-OctoThinker-8B-Base2025.120.13.75
VanillaBackbone=Llama-OctoThinker-3B-Base2025.1201.68