Fronto-Striatal Circuits Optimize Feature-based Attention and Learning

Our new publication (Oemisch et al. (2018) Feature Specific Prediction Errors and Surprise across Macaque Fronto-Striatal Circuits during Attention and Learning) provides the first 4-brain-area survey of how prediction error information in the anterior cingulate – ventral striatum and lateral prefrontal – caudate fronto-striatal loops relate to feature-based attention and learning. We found prediction errors that encode the specific stimulus feature that was reward relevant. This coding took place with stimuli having multiple feature dimensions. Reporting that neurons track the specific reward relevant feature suggests an attractive solution of credit assignment through a distributed feature-specific eligibility trace enabling ‘goal-directed’ synaptic plasticity changes across the entire fronto-striatal network.

Related News

Theta and Beta Frequency

Theta and beta frequency range coherence between anterior cingulate cortex and frontal eye field indexes the successful preparation for anti-saccades and maintenance of working memory content – with larger ACC to FEF direction of granger causal information flow! These important findings is now published in Nature Communications by Sahand Babapoor-Farrokhran and Stefan Everling with contributions […]

First Neuropharmacological Contribution

Congratulations to Ali and Mariann for the first neuropharmacological contribution from our laboratory with the article A computational psychiatry approach identifies how alpha-2A noradrenergic agonist Guanfacine affects feature-based reinforcement learning in the macaque . This study first surveys all 14 different tasks that have ever been used with Guanfacine in nonhuman primate studies and than […]

New 3D-object type with controllable feature space published with open-sourced code

The lab has a new publication showcasing and describing details of Quaddles: A multidimensional 3D object set with parametrically-controlled and customizable features. Quaddles have 5+ feature dimensions, each with multiple possible feature values that can be parametrically morphed, making it possible to generate a near arbitrary number of unique objects. Thanks to Marcus and Milad […]