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

Mathematical Reasoning on DeepMath 2025 (test)

55.8Pass@1

FedAvg-PubSwap

13.26424.30735.3546.393Apr 14, 2026
Updated 3d ago

Evaluation Results

MethodLinks
2026.04
55.8
2026.04
54.6
2026.04
54.5
2026.04
53.8
2026.04
53.3
2026.04
53.1
2026.04
53.1
2026.04
53.1
2026.04
52.7
2026.04
52.6
2026.04
52.3
2026.04
52.3
2026.04
52.1
2026.04
52
2026.04
51
2026.04
50.7
2026.04
50.5
2026.04
50.4
2026.04
49.5
2026.04
49.4
2026.04
49.3
2026.04
48
2026.04
48
2026.04
47.7
2026.04
34.9
2026.04
34.9
2026.04
34.9
2026.04
34.9
2026.04
14.9
2026.04
14.9
2026.04
14.9
2026.04
14.9