1 paper · 1 filter
Albert Alcalde, Leon Bungert, Konstantin Riedl +1
Transformers with self-attention modules as their core components have become an integral architecture in modern large language and foundation models. In this paper, we study the e…