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

Mathematical Reasoning on LMB (LiveMathBench)

35.8Accuracy

RePro

Updated 1mo ago

Evaluation Results

Method	Links
RePro 2025.12		35.8
Original 2025.12		35.3
GRPO 2025.12		35.1
RF++ B + RePro 2025.12		22
PPO 2025.12		20.5
PPO + RePro 2025.12		20.5
GRPO + RePro 2025.12		19.5
PPO 2025.12		19.5
Original 2025.12		18.8
PPO + RePro 2025.12		18.8
GRPO 2025.12		18.8
RF++ B 2025.12		18.5
RF++ B + RePro 2025.12		17.8
RF++ B 2025.12		17.5
GRPO + RePro 2025.12		17.5
GRPO 2025.12		16
PPO + RePro 2025.12		15.8
Original 2025.12		15
RePro 2025.12		14.5
PPO 2025.12		14.5
GRPO 2025.12		12.5
Original 2025.12		10.5
Original 2025.12		7.5