suppose we have a unique floating point data structure with the following layout: first 1 bit for sign bit, next 40 bits for mantissa field, and the remaining 23 bits for the exponent field; what is the proper bias value for the biased exponent?