Salva Ardid

This study examines neural representations in task-switching feedforward networks using context biases, finding that they transition from parallel processing to independent subnetworks, similar to cortical processing.

This study uses biophysically-detailed modeling to show that prefrontal cortex oscillations and local inhibition gate signal propagation from working memory, with resonant frequency tunable by deep layer excitability.

The paper shows that persistent activity in a prefrontal cortex model requires homogeneity in inhibitory neurons and synapses, which can be achieved through homeostatic plasticity.

The paper presents a corticostriatal model showing how biased competition can occur without input bias through physiological asymmetries in striatal neurons.

The paper shows neurons in fronto-striatal circuits encode feature-specific prediction errors during attention and learning, contributing to feature-based attention updates.

This paper shows that prefrontal cortex networks with strong feedback inhibition generate internal oscillations and selectively respond to external oscillations at similar frequencies, with resonance properties that can be tuned.