How Can We Track Cognitive Representations with Deep Neural Networks and Intracranial EEG?

Studies harnessing the heuristic potential of Deep Neural Networks (DNNs) and the high spatiotemporal resolution of intracranial EEG (iEEG) data are providing fine-grained details about the nature of brain representations. Originally employed to simulate visual processing and object recognition in the ventral visual stream (VVS), DNNs have recently been applied as models of brain representation in higher-order cognitive domains. In addition to perception, this approach has recently yielded unprecedented insights in memory research, revealing the representational formats and temporal dynamics of working and episodic memory representations. In this chapter, the potential and the limitations of modeling cognitive representations with DNNs and iEEG data are reviewed.