bfloat-bits>float ( bits -- x )

Vocabulary

math.floats.small

Inputs

bits	an integer

Outputs

x

a float

Word description

Decodes the low 16 bits as BF16. Finite values are exact; NaNs become a canonical quiet NaN.

Definition

USING: math.floats.small.private ;

IN: math.floats.small

: bfloat-bits>float ( bits -- x ) 7 8 small-bits>float ;