Fictive learning signals support flexible social behaviour

  • Datum: 02.09.2019
  • Uhrzeit: 11:15 - 12:15
  • Vortragende(r): Sara Ershadmanesh
  • University of Tehran
  • Ort: Max-Planck-Ring 8
  • Raum: room 203
  • Gastgeber: Peter Dayan
Fictive learning signals support flexible social behaviour

A key feature of healthy social function is the ability to map private states onto public actions in a manner that is appropriate for the current context. For example, it may be appropriate to express gratitude for a gift we do not want, or remorse for an action we do not regret. In this talk, I will present our recent work on how people solve this mapping problem, using decision confidence as a model system for quantifying how private states are formed and how these are mapped onto public actions according to the social context. I will present behavioural and modelling analyses which show that both factual and fictive learning signals underpin the acquisition of context-dependent private-public mappings – where the latter requires imagining what would have happened had another action been taken. I will also present preliminary analyses from an fMRI study in which we sought to identity neural correlates of these learning signals. Finally, I will discuss how our work may shed light on clinical conditions characterized by social dysfunction.

Zur Redakteursansicht