Artem Basharin

Faster Language Models with Better Multi-Token Prediction Using Tensor Decomposition

Black-Box Approximation and Optimization with Hierarchical Tucker Decomposition