bioRxivJClub
Reward changes the use and structure of predictive spatial maps
bioRxiv · · Preprint · Open access
Oza, Cohen + 3 more
Abstract ↗AI summary
The abstract is read at the publisher; the summary is JClub's.
Reward-related changes in hippocampal predictive spatial maps are separable: goal-directed behavior strongly concentrates place fields near reward, while representational reshaping is smaller and more variable.
- Why it matters: It addresses whether reward alters hippocampal place-cell activity mainly by changing how an existing spatial map is used or by reshaping the map through plasticity.
- What they did: A two-pathway model used a recurrent network to learn predictive spatial representations and a reinforcement-learning controller to combine that representation with an explicit goal signal; identical trajectory replay isolated representational plasticity.
- The result: A reward-proximity teaching signal made representational changes larger and more consistent, attenuated when recurrent plasticity was restricted; after reward relocation, directional tuning adapted more fully than place-field organization, and withdrawing reward and proximity teaching reversed field concentration.
The findingWhy it mattersWhat they didThe result
- Open access