Scaling Laws for Neural Language Models (first "scaling laws" paper from 2020) | Dark Hacker News