Contact

Dr. Daniel Braun
Opens window for sending emaildaniel.braun[at]tuebingen.mpg.de

 

Research Assistant in Neurophysics

Thermodynamic Models for Intelligent Systems

Information is a central concept both in physics and in computer science. Intelligent systems that perceive, learn and act are essentially information-processing systems. Physical systems, while usually not thought of as doing information-processing, are also fundamentally constrained by information, as traditionally studied in thermodynamics and statistical physics. In this project we study the hypothesis that thermodynamic laws that constrain physical systems might also be applicable to information-processing systems in general. Recently, we have applied thermodynamics to model decision-making processes in our group. There are two possible student projects available. First, we are interested in developing sampling methods based on free energy differences to solve sequential decision-making problems. Second, we are interested in applying non-equilibrium thermodynamics to our decision-making model to study learning and adaptation.

 

The HiWi position (83 hrs/month) is paid at a rate of €8 per hour. The Emmy-Noether group is funded by an excellence programme of the Deutsche Forschungsgemeinschaft (DFG). The applicants should have a strong mathematical background and have a keen interest in neuroscience and artificial intelligence. Candidates should send a CV and a brief statement of interest to Dr. Daniel A. Braun Opens window for sending emaildaniel.braun[at]tuebingen.mpg.de

Last updated: Thursday, 09.01.2014