here you goThings I’ve made, read and saved — shared here.

Untitled (https://arxiv.org/pdf/2005.14165.pdf)

You: "🧐Ah mon cheri I have framed this as a very clever convex optimization problem. I'll derive recoverability bounds and include some non-reproducible tables in my paper"

The guy she said not to worry about "Gon' train this 175 billion parameter model"