Seminars: D. A. Rybin, Acceleration of $XX^{T}$ and Machine Learning for Mathematics

SEMINARS


Mathematical Foundations of Artificial Intelligence April 1, 2026 17:00, Moscow, Steklov Mathematical Institute + Kontur Talk

Acceleration of $XX^{T}$ and Machine Learning for Mathematics D. A. Rybin^ab ^a Chinese University of Hong Kong ^b Autokernel.cn
https://vkvideo.ru/video-222947497_456239185 https://youtu.be/1kIFJkl-mic Abstract: In this talk, we will examine examples of using machine learning to search for mathematical objects. In particular, we will present our result regarding the first acceleration of the $XX^{T}$ matrix multiplication algorithm in 50 years. We will also discuss cutting-edge results in the application of large language models to purely mathematical problems.