This is the website for the paper “Motifs in Attention Patterns of Large Language Models”, which aims to provide a new perspective on categorization of attention heads in LLMs by studying the structure of attention patterns.