Hacker News new | ask | show | jobs
by Grosvenor 612 days ago
https://news.ycombinator.com/item?id=36871528

Hah. Yes. It looks like they only show up in models with 6.7B parameters or more.

The problem can start at 125M. Small enough to test on a whim.

So train a model that exhibits these behaviours, then try it out.