Here's a very interesting paper from the #KeinathLab:
In the absence of task or reward, #PlaceCells appear to reflect the past history of behavioural biases, i.e. the accumulated past paths taken by each individual mouse. I am actually starting to think that #HippocampalReplay is doing a similar thing - just reflecting an average of the individual's past behaviour, and that all supposed "coding" of reward is purely driven by this (notice how reward "coding" always seems to happen in cases where reward biases the behavioural patterns..) 馃憖馃馃