Hippocampal hierarchical networks for space, time, and memory

- The hippocampus represents the local environment as well as experiences embedded in it.
- Key hippocampal mechanisms are relational processing, multi-scale representations and remapping.
- In combination, these computational mechanisms enable formation of hierarchical networks in memory.
- These coding principles are domain general rather than domain specific.

The hippocampus is critically involved in both mapping space as well as the formation of memories for events. Here, we propose that common coding principles in the hippocampus enable spatial, temporal and episodic representations. We discuss recent studies employing novel cognitive tasks as well as newly developed representational analysis techniques which show that both spatial and mnemonic representations can be thought of as networks of interlinked elements, be it locations in space or events in memory. These mnemonic networks share certain characteristics, such as plasticity and hierarchical organisation, which enable structured representation of information while also allowing simultaneous assimilation of new elements. We conclude by outlining possibilities of how neural mechanisms underlying the formation of such networked representations can support the organisation of interlinked information beyond time and space.