Data Elixir on Mythryl: “Why does neural network training almost never fail? Pure…”

Why does neural network training almost never fail? Pure combinatorics. A 6K parameter network contains 10^1089 possible sparse subnetworks. That's 10^900 solutions per atom in the universe. We're not smart, we're just brute forcing.

embedding-space.github.ioSparse Networks and Lottery WinnersEmbedding Space is a blog about machine learning and artificial intelligence.

24/10/25, 2:47 pm0 reposts4 likes

Post