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

Math on GSM8K (test)

94.04Mean@4

TextGrad-M

-0.641623.939248.5273.1008May 31, 2025Jul 17, 2025Sep 3, 2025Oct 21, 2025Dec 7, 2025Jan 24, 2026Mar 13, 2026
Updated 25d ago

Evaluation Results

MethodLinks
2025.05
94.04-74.121.73
2025.05
93.98-74.592.2
2025.05
93.74-72.39-
2025.05
93.15-72.96-
2025.05
91.55-72.11-
2026.02
91.4---
2025.05
90.78-72.630.8
2025.05
89-71.83-
2025.05
88.74-68.790.94
2025.05
87.99-67.84-
2025.05
86.9-68.91.05
2026.02
82.3---
2026.02
70.7---
2026.03
54.353.8--
2026.03
50.949--
2026.03
48.449.3--
2026.03
47.648.4--
2026.03
4348.9--
2026.03
42.240.4--
2026.03
39.338.6--
2026.03
38.952--
2026.03
36.849.8--
2026.03
34.345.8--
2026.03
33.750.9--
2026.03
33.351.4--
2026.03
32.848.4--
2026.02
22.3---
2026.03
326.2--