New FNN SoTA, Fake Turing Test & Mixed Learning Rates

Hey, sorry for putting 3 topics in the same thread but I’m not sure how the trust system works on here & don’t want to get stopped halfway through sharing stuff.

Anyway, here’s what I’ve been up to over the last two weeks:

20% accuracy boost on SNN’s by introducing some superstructures: YouTube / GitHub

Learning rate mixtures (sliced vertically instead of horizontally): YouTube / Gist

Fake Turing Test dataset: Kaggle

1 Like