Dear Authors,
Your study on AI-driven predictions of mathematical problem-solving beliefs is intriguing, but several aspects merit further clarification. Could you provide more details on how overfitting was addressed in your ANFIS and ANN models, and discuss the implications of moderate correlations (r = 0.564 for ANFIS, r = 0.422 for ANN) on the predictive validity of your findings? Additionally, the theoretical link between creative thinking dispositions and problem-solving beliefs needs more robust justification, along with a critical discussion of potential biases introduced by the convenience sampling method. Clarifying these points would strengthen the impact and reliability of your work.