When working with scientific notation, what does a negative exponent let you know about the size of the number?