View job here

Lead PI:

Max Wolf, Jens Krause

Description of the postdoctoral project

This project addresses one of the most fundamental questions associated with exploration behavior and the development of intelligent behavior in biological and artificial systems. When confronted with an unknown environment, what are the algorithmic rules underlying the development of intelligent (i.e. adaptable, general, cost effective and goal-directed)  behavior? In order to address this question, the postdoctoral researcher will develop and test algorithmic models of exploration behavior that (i) predict real world behavioral-experiential trajectories of newborn individuals (see below) and (ii) perform well when implemented in agents in virtual experiments simulating a broad range of conditions and tasks. Throughout, the postdoctoral researcher is expected to closely collaborate with a second postdoc, who focusses on mapping the exact behavioral-experiential trajectories of genetically identical fish individuals from day 1 of their life.


Applicants must hold a PhD in Cognitive Science, Biology or related fields. Applicants should have strong quantitative skills and a firm background in several of the following topics:

  • Lab-based experiments with fish
  • Analysis of high-resolution tracking-data
  • Machine learning
  • Agent-based computer simulations
  • Mathematical modelling of animal and/or human behavior

Excellent English skills both written and spoken are required.

Application deadline: 29 November 2019 23:59 CET

Please visit the website of the Humboldt University, which gives you access to the legally binding German version.