abstract Carlo Reverberi

Carlo Reverberi (Psychology department, Università Milano Bicocca, Milano, Milan Center for Neuroscience, Milano)

Representation and discovery of task features in human frontal cortex

Humans are able to flexibly devise and implement rules to reach their desired goals. In some situations, we can use simple rules, such as “if traffic light is green then cross the street.” In most cases, however, more complex rule sets are required, involving the integration of multiple layers of control. We explored the neural representation of rule sets in a series of fMRI experiments using multivariate pattern analysis. We found that the neural representation of complex rules is “compositional”, i.e. built on the neural representation of their constituent rules or features. Complex rules are “decomposed” in their elementary features. The constituent features are then represented in different brain structures, depending on the type of information to be stored.

In particular, we found that right ventro-lateral prefrontal cortex (BA47) consistently encoded the relevant conditional (if-then) rules. By contrast, pre-motor cortex (BA6) encoded the order in which rules should be applied. Left inferior frontal gyrus (BA44) is involved in the encoding of the logical relations holding between the constituents of complex rules. Finally we found that medial prefrontal cortex encodes features relevant for new, yet to be implemented, rule sets. We argue that compositionality, “decomposition” and information-guided representation are general features of prefrontal cortex functional organization.