Mnemonic prediction errors bias hippocampal states

In situations when our experience violates our predictions, it is adaptive to upregulate encoding of novel information, while down-weighting retrieval of erroneous memory predictions to promote an updated representation of the world. We asked whether mnemonic prediction errors promote distinct hippocampal processing states by leveraging recent results showing that encoding and retrieval processes are supported by distinct patterns of connectivity, or states, across hippocampal subfields. During fMRI scanning, participants were cued to retrieve well-learned room-images and were then presented with either an image identical to the learned room or a modified version (1-4 changes). We found that CA1-entorhinal connectivity increased, and CA1-CA3 connectivity decreased, with the number of changes to the learned rooms. Further, stronger memory predictions measured in CA1 during the cue correlated with the CA1-entorhinal connectivity increase in response to violations. Our findings provide a mechanism by which mnemonic prediction errors may drive memory updating - by biasing hippocampal states.