A mathematical perspective on Transformers
2023-12-17Code Available1· sign in to hype
Borjan Geshkovski, Cyril Letrouit, Yury Polyanskiy, Philippe Rigollet
Code Available — Be the first to reproduce this paper.
ReproduceCode
- github.com/borjang/2023-transformers-rotfOfficialIn papernone★ 39
Abstract
Transformers play a central role in the inner workings of large language models. We develop a mathematical framework for analyzing Transformers based on their interpretation as interacting particle systems, which reveals that clusters emerge in long time. Our study explores the underlying theory and offers new perspectives for mathematicians as well as computer scientists.