1 paper
Wei Shi, Shuang Li, Kerun Yu +9
There is a growing interest in expanding the input capacity of language models (LMs) across various domains. However, simply increasing the context window does not guarantee robust…