undefined | Better HN

0 pointsamelius1y ago0 comments

You might then as well replace the multiplication by the addition in the original network. In that case you're not even approximating anything.

Am I missing something?

0 comments

1 comments · 1 top-level

dotnet001y ago

They're applying that simplification to the exponent bits of an 8 bit float. The range is so small that the approximation to multiplication is going to be pretty close.

j / k navigate · click thread line to collapse