List your company

Mathematical Reasoning on MATH-500 2023

13.6Peak AccuracyPessimism Only
11.41611.98312.5513.117Apr 6, 2026

Evaluation results

MethodLinks
Pessimism Only2026.0413.611.91.7
RM + Pessimism2026.0412.310.12.3
Reward Model2026.0411.58.53