Generating the significand :

  When specifying a decimal value in scientific notation :

    6.02  x 10^23
 
  Most significant digit to the of the decimal with value of 1-9

  Because it value is variable, it must be stated.

  
  However, in binary, the most significant digit can only be 1

  Our number : 5.75 or 101.11b

    101.11 =  1.0111 * 2^2 (unbiased mixed base example)

    2^(3-1)-1 = 3 bias     3 (bias) + 2 exponent = 5 or 101b
       ^
      3 bit exponent
                     coefficient   exponent (now bias and binary)
  Binary   101.11  = 1.0111     x   101 

  Because a binary expressed in scientific notation will always have
  a 1 in the integer position,

    We can assume the one is there and don't have to store it.

    This gives us the ability to store one additional bit of precision.

    We only need to store 0111 because we know the most significant digit
      is 1, this modified coefficent is called the significand.
      The "missing" 1 is called the hidden bit.

  If your number has more significant digits than you have room to store,
  then unstorable less significant bits are discarded. 

  Depending on the protocol, some form of rounding may occur. 

  Examples