Im new to Machine Learning, so please just dont blast me.

I studied that L2 Regularization in Logistic regression creates a sort of sphere that limits the choice of w weight ( info Here )

But, mathematically why this happens ? Thanks