Philip Resnik

The paper uses MEG to show a hierarchical progression of speech processing from acoustic to linguistic features, with late responses indicating predictive coding and semantic processing.

The paper introduces Eelbrain, a Python toolkit for time-continuous analysis using temporal response functions, demonstrating its application to EEG data from audiobook listening.

This study investigates how the brain processes speech by using both local and unified predictive models, showing that listeners employ these models in parallel to anticipate phonemes.