Computational Neuroscience InitiativeBarchi Library - 140 John Morgan Building
Alexandre Filipowicz
Kable Lab, University of Pennsylvania
Adapting to a changing world
To make sense of the world, humans build beliefs from past observations to guide actions and expectations. In changing environments, effective beliefs need to balance being both flexible and adaptive to new information, while also being resistant to noisy, spurious events. In this talk I will be presenting research aimed at characterizing the processes that tradeoff robustness to noise and flexibility to change in belief updating. I will first present results from a study that measured the systematic relationship between unexpected, 'surprising' events and their influence on belief updating. I will then present the concept of 'complexity', a general property of beliefs that guides flexibility in learning. Using this framework I will present a study that use complexity to characterize individual differences in belief updating, and another that uses complexity to explain disruptions in belief updating that follow certain forms of brain injury. The talk will finish by discussing some of my current and future work on using the concept of complexity to provide a better understanding of the information processing tradeoffs that guide human learning and decision making.
This will be a practice job talk. A pizza lunch will be served.