Why do I get: -9223372036854775808

The one glaring error is this:

int64_t maxPower = 50;
int64_t results[maxPower];  // <-- Not really legal, but let's pretend
//...
for (int64_t power = 0; 
             power <= maxPower; // <-- Look at the loop constraints
             power++) 
{  
    //....
}
results[power] = value;  // <-- Buffer overrun

Since maxPower is 50, you are accessing an element that’s out of bounds. This is undefined behavior

Also, to make the code legal C++, either declare an array of 50:

int64_t results[50];

or declare a std::vector<int64_t>.

When you do that and use at() instead of [], then you get a better view of how your program breaks. See the Live Example using std::vector here.

---

As to the “odd number”, you are getting a numeric overflow here:

int64_t value = pow(tuple, power);

If tuple is 3, and power is 41, then the value overflows the int64_t type.

Maybe you need to use an arbitrary precision library (one that preferably has overloaded the requisite mathematical operators) or write your own, instead of using the “regular” C++ types. Then you won’t or shouldn’t overflow.