Dimitrije Marković

Postdoc at TU Dresden, Chair of Neuroimaging.

As a trained theoretical physicist, I have a deep interest in computational principles that offer unifying description of various cognitive phenomena, neuronal dynamics, and human and animal behavior.

I currently focus on investigating computational mechanisms that underlie learning, inference and planning in the presence of various sources of uncertainty. I describe these processes in the context of Bayesian inference, planning as inference, and active inference.

Recently, I started investigating how the representation of time (or event durations) is used for planning and decision making in dynamic environments. An interesting question here is how temporal representations modulate the decision-making process and under which conditions this interaction might lead to ‘suboptimal’ decisions.

On the long run, I am interested in connecting the normative models of human behavior to neuronal level and self-organizing plasticity mechanisms that shape neuronal dynamics.

I am currently employed in the SFB 940, where I work on models of cognitive control and goal-directed actions, and apply these models to understand human behavior in specific behavioral experiments. Here is an up-to-date curriculum vitae.