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

Step-level correctness assessment on GSM8k (test)

70.2PR-AUC

Qwen2.5-Math-PRM-7B

3.32820.68938.0555.411Nov 9, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.11
70.2
2025.11
67.3
2025.11
65
2025.11
62.4
2025.11
61.6
2025.11
61.3
2025.11
59.4
2025.11
54
2025.11
53.7
2025.11
41.2
2025.11
40.5
2025.11
17.1
2025.11
12.2
2025.11
11.9
2025.11
11.6
2025.11
10.6
2025.11
10.1
2025.11
9.9
2025.11
9
2025.11
8.9
2025.11
6.1
2025.11
5.9