A computational problem is defined by specifying how the output relates (mathematically) to the output.
Example: modular inverse problem
input: modulus m, standard name b
output: standard name c such that 
Example: modular cube problem
input: modulus m, standard name b
output: standard name c such that 
Example: modular exponentiation problem
input: modulus m, standard name b, positive integer k
output: standard name c such that
An instance of a computational problem is an assignment of numbers to the inputs to the problem.
Example: an instance of the modular inverse problem is ``the inverse of 160 modulo 937971''