Zobrazeno 1 - 3
of 3
pro vyhledávání: '"Basharin, Artem"'
We propose a new model for multi-token prediction in transformers, aiming to enhance sampling efficiency without compromising accuracy. Motivated by recent work that predicts the probabilities of subsequent tokens using multiple heads, we connect thi
Externí odkaz:
http://arxiv.org/abs/2410.17765
We develop a new method HTBB for the multidimensional black-box approximation and gradient-free optimization, which is based on the low-rank hierarchical Tucker decomposition with the use of the MaxVol indices selection procedure. Numerical experimen
Externí odkaz:
http://arxiv.org/abs/2402.02890
Autor:
Bakinova, Ekaterina, Basharin, Artem, Batmanov, Igor, Lyubort, Konstantin, Okhotin, Alexander, Sazhneva, Elizaveta
Publikováno v:
In Information and Computation February 2022 283