Representational drift for memory and perception

• Applicant: Professor Dr. Raoul-Martin Memmesheimer
• Subject Area: Experimental and Theoretical Network Neuroscience
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 564152855

Organisms change over time, in many ways and on different levels. Some of these changes are well known from daily life; the knowledge about others is the surprising result of recent research. An example for the latter is that the neuronal structures that underlie memories, the representations, gradually change over time. The behavior and therefore perhaps also the subjective percept of the memory, can nevertheless remain unchanged. In the proposed project we will create models at different levels of abstraction that explain this so-called representational drift by the random joining and leaving of neurons. More abstract models will describe the drift through the entire brain. They might lead to an explanation of experimentally found memory consolidation by a completely random process. In a second part, we aim to develop biologically plausible models for specific types of representations, which underlie for example gustatory sensations. Their long-term changes have not yet been measured. Our research shall show whether we expect them to drift as well. Finally using experimental data we will test our hypothesis that representational drift can be compensated by intrinsic adaptive processes that do not require external feedback. For this, we will especially use neurobiological data that have been or are currently acquired by experimentally working collaborative partners.

• DFG Programme: Research Grants