Polina Tabakova decided to apply for a Philology degree at HSE in Nizhny Novgorod because she grew up in Mari El and did not want to move far away from the Russian forests. In an interview for the Young Scientists of HSE University project, she spoke about the genre of the campus novel, the existential drama of Kolobok, and a blackout version of Eugene Onegin.
Researchers from the AI and Digital Science Institute at the HSE Faculty of Computer Science have developed a new compression method for large language models such as GPT and LLaMA that reduces their size by 25–36% without additional training or significant loss of accuracy. This is the first approach to use mathematical transformations—specifically, rotations of model weights—to make models more amenable to compression with structured matrices. The study results have been published in ACL Findings 2025. The code is available on GitHub.
Malakhov S., Rozhkov M., Prikladnaya Diskretnaya Matematika 2022 No. 56 P. 17–27
The objective of this paper is to suggest a method of the construction of circulant matrices, which are appropriate for being MDS (Maximum Distance Separable) matrices utilising in cryptography. Thus, we focus on designing so-called bi-regular circulant matrices, and furthermore, impose additional restraints on matrices in order that they have the maximal number of some ...
Rozhkov M., Malakhov S., Journal of Applied and Industrial Mathematics (перевод журналов "Сибирский журнал индустриальной математики" и "Дискретный анализ и исследование операций") 2019 Vol. 13 No. 2 P. 302–309
MDS matrices are widely used as a diffusion primitive in the construction of block type encryption algorithms and hash functions (such as AES and GOST 34.12–2015). The matrices with the maximum number of 1s and minimum number of different elements are important for more efficient realizations of the matrix-vector multiplication. The article presents a new method for the MDS ...
Malakhov S., В кн.: Межвузовская научно-техническая конференция студентов, аспирантов и молодых специалистов им. Е.В. Арменского.: МИЭМ НИУ ВШЭ, 2018. С. 218–220.
Нижеследующая работа посвящена экспериментальному оцениванию числа MDS матриц специального вида над конечными полями. Оценки получены на основании серии экспериментальных построений MDS матриц с помощью ПЭВМ и системы компьютерной алгебры Wolfarm Mathematica. ...