I need to make sure you undertand this part: Floats are scientific/exponential notation. The memory reserved for a float contains the mantisse (signiticant digits) and the exponent over 10. That exponent is how the floating (moveable) part of floating point works. That is how they cover value from "size of the universe" to "size of the atom" ranges.
You can represent them as a human reable number. But doing so needs so insanely many zeroes, most people do not bother. 45-38 zeroes for float in fact. And it will not get you any additonal data or precision, as the math is still done by the CPU in float notation. Writing them out with Zeroes is just a pointless waste of characters (and thus bytes), so outside of input from or output towards the user, nobody will bother.
Not that I think 150,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000 is really that human readable to begin with. Honestly 1.5 * 10^45 seems way more readable for me. And everyone that else would ever use such a number.