Your question is Mitigate Model Overfitting. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You've trained a model that looks strong on the training set, but validation performance is noticeably worse. The team wants to know how you would reduce the gap without guessing.
How do you mitigate overfitting?