[mikequinlan]

This can (obviously) be done to replace any multiplication. The problem is that 2 logs, an add, and an exponentiation are a lot slower than a single multiply.

[danbst]

The idea is to store all matrix weights for neural model in log space, and never leave it. Maybe create a new activation function which doesn't leave log space.

You can check the code, I don't just replace multiplication, I perform addition (with tricks) in log space

[BoiledCabbage]

Could it (ML learning) all be done in "log-space"? Ie never need to pay the cost of the initial transformation and return?