Mikhail Dmitrievich Lebedev

This paper introduces DualMyo, a Transformer-based model for EMG-to-digit classification, achieving 91% accuracy with lightweight fine-tuning for cross-session adaptation.

This paper uses large language models to improve P300 speller output by reducing errors from fewer stimulus repetitions, enhancing communication efficiency.

This paper compares data aggregation strategies for improving P300 classification using EEG, finding that controlled cross-subject averaging with multi-channel models achieves high Information Transfer Rate.

The paper investigates relative-location tuning in five frontal areas of macaque monkeys, finding that neurons encode spatial information based on relative object positions rather than absolute coordinates.

This paper reanalyzes Hafting et al.'s 2005 grid cell data, clarifying that only one grid axis is tilted between subpopulations and that navigation traces align with grid orientation.