David Richter

Attention and Explicit Knowledge Drive Predictive Sharpening in Early Visual Cortex

The paper investigates how attention and explicit knowledge influence predictive sharpening in early visual cortex using fMRI.

Rapid Computation of High-Level Visual Surprise

The paper uses EEG and deep neural networks to show that high-level visual predictions, not low-level features, drive neural responses to unexpected stimuli around 200ms after stimulus onset.

Abstract rule generalization for composing novel meaning recruits the frontoparietal control network

The study investigates how the brain generalizes abstract linguistic rules to form novel meanings, finding that a frontoparietal network, not the medial prefrontal-hippocampal network, supports this process.

High-level prediction errors in low-level visual cortex

The paper investigates prediction errors in visual cortex using fMRI and deep learning metrics, finding that high-level predictions influence early visual areas.

Proactive distractor suppression in early visual cortex

The paper shows that early visual cortex suppresses potential distractors based on learned spatial priors, even before stimuli are presented.

An adversarial collaboration to critically evaluate theories of consciousness

A preregistered multimodal neuroimaging study across 256 human participants directly tests Integrated Information Theory against Global Neuronal Workspace Theory, confirming some predictions while revealing major empirical challenges for both.

Updating contextual sensory expectations for adaptive behaviour

This study shows that contextual sensory expectations update in a hierarchical manner, with executive control regions showing suppression first, followed by sensory areas, when expectations are task-relevant.

Conceptual associations generate sensory predictions

The paper shows that conceptual associations generate sensory predictions, suppressing neural responses to expected visual stimuli.

Spatial and temporal context jointly modulate the sensory response within the ventral visual stream

This study investigates how spatial and temporal context jointly modulate neural responses in the ventral visual stream using fMRI, finding that spatial context has a stronger modulatory effect than temporal context.

Statistical learning attenuates visual activity only for attended stimuli

The paper shows that statistical learning reduces neural activity for expected stimuli, but only when those stimuli are attended.

Suppressed sensory response to predictable object stimuli throughout the ventral visual stream

The paper investigates expectation suppression in the ventral visual stream using fMRI, finding robust suppression across areas including V1, LOC, and anterior visual areas, with dampening in object-selective regions.