AI RESEARCH
LLMs learn backwards, and the scaling hypothesis is bounded. [D]
r/MachineLearning
•
Submitted by /u/preyney [link] [comments]