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

Mathematical Reasoning on Minervamath

34.19Accuracy

GRPO + ARROL

2.074810.412418.7527.0876Mar 25, 2026Apr 4, 2026Apr 15, 2026Apr 26, 2026May 6, 2026May 17, 2026May 28, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.03
34.191.6259.53
2026.03
32.36-56.66
2026.05
31.3--
2026.03
30.88-51.01
2026.05
30.1--
2026.03
28.671.6353.54
2026.05
28.6--
2026.03
20.96--
2026.03
20.961.7-
2026.03
17.65-34.79
2026.03
16.911.6137.09
2026.03
4.041.6717.49
2026.03
3.31-14.63