Timothée Proix

The paper shows top-down semantic predictions align phonetic neuronal dynamics in the human superior temporal gyrus during speech processing.

The paper proposes a computational framework where cortico-subcortical circuits implement multi-head self-attention, aligning with human brain recordings during speech perception.

The paper investigates hierarchical syntactic planning in speech production using SEEG recordings, finding that core syntactic categories are planned before modifiers and that the hippocampus contributes to sentence structure building.

The paper shows that human medial temporal lobe neurons encode letter pairs through a segregated population during working memory, independent of single letter encoding.

The paper maps directed cortico-limbic signaling in the human brain using electrical probing, finding limbic structures send twice as many signals as they receive during wakefulness and sleep.

The paper introduces a diffusion-approximation framework to model and interpret neuronal spiking dynamics using nonlinear Hawkes processes, showing its accuracy and applicability to human and nonhuman primate data.

The paper shows that oscillatory trajectories in dimensionality reduction of neural activity data are not evidence of latent dynamics but result from continuous temporal variation in the data.

The paper decodes imagined speech using low- and cross-frequency neural features from electrocorticography data, showing their effectiveness in perceptual speech spaces.

This paper demonstrates that intracortical signals from areas distant from seizure onset can predict focal seizures with 90% accuracy using machine learning.