Modeling Autonomous Shifts Between Focus State and Mind-Wandering Using a Predictive-Coding-Inspired Variational RNN Model
The episode explores a paper modeling autonomous shifts between focus and mind-wandering using a predictive-coding model. The hosts discuss how these attention shifts are not conscious decisions but are driven by a dynamic mechanism involving the parameter 'w'. When prediction error is low, 'w' increases, causing the system to rely on internal thought (mind-wandering). This provides a quantifiable framework for understanding cognitive transitions.