I need help with understanding multiple linear regression for my exam coming up. So, I know the model is

y=B0 +B1x1+...+Bkxk+e

(e = episolon for error)

Is there a scenario where the regressors may change the value of E(y|x1, x2) differentially as the value of the other regressor changes (say we're given x1 and x2 as non-random regressors)? How would I be able to assess the potential change statistically? I don't understand