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

Numerical Reasoning on TaTQA

92.9Accuracy

GPT-5 mini-high

26.8644.00561.1578.295Mar 7, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.03
92.977.1
2026.03
9067.7
2026.03
89.374.6
2026.03
87.672.5
2026.03
87.171.5
2026.03
8772.1
2026.03
85.174.7
2026.03
8270.3
2026.03
81.568
2026.03
8061.4
2026.03
78.347.8
2026.03
69.356.9
2026.03
49.336.7
2026.03
29.436.2