alt Hacker News~85% accuracy on MNIST. Sigh.
How does it do on CIFAR-10, or even better, ImageNet?
Interesting research, not sure it's a backprop alternative.
===
EDIT: accuracy on MNIST is not ~90%. It's ~85%.
Replies
qarl • today at 12:06 AM
They state replacing backprop is not their goal.
Their goal is to understand how distributed systems which cannot do backprop (the brain) can still do learning.
➕ show 1 reply
bz_bz_bz • yesterday at 10:13 PM
Their image classification benchmarks include both: https://pub.sakana.ai/pc-alm/assets/figures/benchmark_accura...
➕ show 1 reply
It might be beneficial while not being optimal on its own.
The obvious example is if it has different behaviour around local minima, it could be an altenate pathway out.
I have often wondered if doing training with radically different aproaches for the first few iterarions would avoid any method specific artifacts before the weights had time to denoise.