Hacker News new | ask | show | jobs
by danielhanchen 324 days ago
I wrote approximately in the blog about it and linked some papers! I also wrote about it here - https://unsloth.ai/blog/dynamic-4bit - one has to inspect the activation and weight quantization errors!
1 comments

So you are basically looking at "fMRI" of the "brain" while it's doing a wide range of things and cutting out the things that stay dark the most?
Oh that's a good analogy! Yes that sounds right!