Discussion about this post

User's avatar
Sebastian Sartor's avatar

Just learned that Bell Labs already found neural scaling laws in 1993 and that scaling laws have a much older history than I thought. Here's the link to the NeurIPS paper: https://proceedings.neurips.cc/paper/1993/file/1aa48fc4880bb0c9b8a3bf979d3b917e-Paper.pdf

Expand full comment
Dominique Paul's avatar

Hey, cool post! If I only read one of the papers analyzing scaling laws for robotics, which one should I read? Also, you look at the reduction in loss for the scaling laws. Does a 50% reduction in loss necessarily also transfer to the same improvement in accuracy/success rate in the respective domains? That would be the obvious metric that matters.

Expand full comment
1 more comment...

No posts