#32 · Systems biology

Recurrent CA3 dynamics may yield both place and time cell representations

Across these items, hippocampal place and time coding are treated as outcomes of how representations are formed and how learning statistics shape them. A recurrent CA3 autoencoder produces both place and time cell representations, and the preprint proposes a shared origin with differences attributed to task-driven factors. A PLOS Computational Biology model integrates environmental-feature and behavioural-outcome inference, overcoming the limits of each alone while reproducing PFC-hippocampus behavioural and cellular phenomena. A second PLOS Computational Biology study applies the deep successor representation to spatial navigation, linking pre-exposure exploration aligned with reward location to later latent learning. What is new is explicit modelling of how content and statistics jointly constrain hippocampal and prefrontal activity, rather than treating place and time coding as separate phenomena.