Which statement about the cumulative distribution function (CDF) F(x) of a continuous random variable is always true?

F(x) must be symmetric about its mean and this symmetry is necessary for F to be nondecreasing and bounded between 0 and 1

F(x) can take values greater than 1 for x near positive infinity if the PDF has unbounded peaks, because the integral of unbounded functions can exceed 1

F(x) is always differentiable everywhere and F'(x) equals f(x) even where the PDF has jump discontinuities or is undefined

F(x) is nondecreasing and 0 ≤ F(x) ≤ 1 for every real x