Comment by zeeshana07x

The gap between how this is described in the paper vs the blog post is pretty wide. Would be nice to see more accessible writing from research teams — not everyone reading is a ML engineer