Figure 2. (IMAGE)
Caption
MathEval evaluation results. (a) Overall average for different model categories; (b) model performance by parameter size; (c) improvements on math domain models; (d) comparison of solving arithmetic and math word problems capabilities of models. (a), (b), and (c) Show the discovery of closed-source models, open-source models, and math domain models. (d) Compares the model-level capabilities across problem type dimensions.
Credit
HIGHER EDUCATON PRESS
Usage Restrictions
none
License
Original content