List your company

Mathematical Reasoning on AIME 24 (Score)

3.33ScoreTD-MNPO
-0.13320.76591.6652.5641Sep 27, 2025

Evaluation results

MethodLinks
TD-MNPO2025.093.33
HT-MNPOReward Model=ArmoRM-Llama32025.093.33
SFT Model2025.090
DPO2025.090
SimPO2025.090
SPPO2025.090
INPO2025.090
HT-MNPOReward Model=Skywork-Reward-V22025.090
HT-MNPOReward Model=Athene-RM-8B2025.090