I am running into a minor issue while using the math/big big integer library. According to the documentation, it should work perfectly fine with negative exponents so long as the modulus is non-zero. However, when running the following piece of code, I get a printed result of 1, instead of the expected result 16. Am I misunderstanding the documentation or doing something wrong? I am using go version 1.10.3 linux/amd64.
package main
import "math/big"
import "fmt"
func main() {
fmt.Printf("%v\n", new(big.Int).Exp(big.NewInt(28), big.NewInt(-3), big.NewInt(47)))
}