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

Step-wise Verification on ProcessBench Math

69.5TNR

Cons@4

12.327.154256.85May 20, 2026
Updated 13d ago

Evaluation Results

MethodLinks
2026.05
69.586.377
2026.05
6985.576.4
2026.05
68.786.976.7
2026.05
68.586.976.6
2026.05
66.88574.8
2026.05
58.48368.6
2026.05
4694.361.8
2026.05
45.891.461
2026.05
21.761.632.1
2026.05
21.569.232.9
2026.05
20.369.431.4
2026.05
1858.427.5
2026.05
14.584.224.7