1 paper
Huanyu Liu, Ge Li, Yihong Dong +7
Large language models (LLMs) based on the Transformer have demonstrated strong performance across diverse tasks. However, current models still exhibit substantial limitations in ou…