Speaker
Dr
Evgeny Sobko
(London Institute for Mathematical Sciences)
Description
In my talk, I will introduce a novel AI-based approach to Integrable Models. I will demonstrate how neural networks can be employed to numerically solve the Yang-Baxter equation and discover new integrable spin-chains. The Hamiltonians of these spin-chains form projective varieties, and I will show how, by using the Boost operator construction for conserved charges, we derive their analytical forms from the approximate numerical data obtained by the neural network. Finally, I will briefly discuss the extension of our method to 2D Integrable Quantum Field Theories.
Author
Dr
Evgeny Sobko
(London Institute for Mathematical Sciences)
