Deciphering the neuronal codes for convergently evolved working memory functions in humans, monkeys, and crows

Andreas Nieder - University of Tübingen
Florian Mormann - University of Bonn Medical Center

Working memory, the ability to briefly retain and manipulate information in mind over seconds to minutes, is the fundamental basis of cognition in humans and intelligent animals alike. Traditional research examining individual neurons primarily in nonhuman primates suggested that information is stored in working memory in ongoing sustained activity; here, neurons maintain elevated firing rates as a static code for working memory over brief periods of time in delayed response tasks. In contrast, more recent recordings have emphasized that information is often stored in working memory in a dynamic population code, where different neurons contain information at distinct intervals during a memory delay. Currently, the functional significance of these putative codes, their generalizability to different phylogenetic taxa, as well as their computational realization in neuronal circuits is unknown. This project uses a comparative approach in order to decipher the neuronal codes and overarching computational principles of working memory in convergently evolved associative endbrain structures of humans, monkeys and crows. For optimal comparability of neuronal data, all three species will be engaged in the same controlled and demanding delayed response task in which working memory will be challenged by distractors. Using intracranial electrophysiological recordings in combination with neural network modeling, the working memory codes that proved to be superior in vertebrate evolution irrespective of the precise anatomical endbrain structures can be identified and theoretically grasped.