1 paper
Evan Wang, Federico Cassano, Catherine Wu +7
While scaling training compute has led to remarkable improvements in large language models (LLMs), scaling inference compute has not yet yielded analogous gains. We hypothesize tha…