Christopher Z Ye

Jun 4, 2024 | arxiv.org | Daniel Kane |Sihan Liu |Christopher Z Ye |Russell Impagliazzo

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Feb 12, 2024 | arxiv.org | Christopher Z Ye

[Submitted on 12 Feb 2024] Title:The I/O Complexity of Attention, or How Optimal is Flash Attention? Download a PDF of the paper titled The I/O Complexity of Attention, or How Optimal is Flash Attention?, by Barna Saha and 1 other authors Download PDF Abstract:Self-attention is at the heart of the popular Transformer architecture, yet suffers from quadratic time and memory complexity.

Sep 23, 2023 | arxiv.org | Christopher Z Ye

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.