#include "tensorstore/util/bfloat16.h"
BFloat16 tensorstore::expm1(BFloat16 x);

Computes e raised to the given power, minus 1 (ex−1).