Hannes Ruge

This study examines thalamic connectivity changes during initial learning of goal-directed behavior using fMRI, finding specific patterns of functional connectivity alterations in higher order thalamic nuclei.

This study uses advanced deep neural networks to predict brain activity during time-continuous stimulus-response tasks, showing that more advanced DNNs provide better predictive features and align more closely with brain organization.

This study uses advanced deep reinforcement learning models to predict human motor responses in time-continuous tasks, finding that more complex models achieve higher prediction accuracy.

The paper examines how functional connectivity between thalamic nuclei and cortical/subcortical networks changes during learning in humans, identifying shifts associated with transitioning from controlled to habitual behavior.

This study uses fMRI to compare neural changes during instruction-based and trial-and-error learning, finding that S-R rule representations guide task performance regardless of learning mode.

The paper provides optimized implementations of L1-regularized models for fMRI analysis, significantly reducing computation time for whole-brain data.

The paper investigates how the brain represents new rules after first-time instruction, finding that the ventrolateral prefrontal cortex (VLPFC) encodes rule-specific states immediately after instruction and stabilizes during early implementation.