Distributed cell assemblies spanning prefrontal cortex and striatum Highlights d Neurons fire at high synchrony (10 ms) in prefrontal-striatal assemblies d For assembly activations, members phase lock to 4 and 8 Hz rhythms d Behavioral correlates are more robustly expressed by assemblies than their members d The assemblies are endogenously generated, spontaneously reactivating during rest

Cell assemblies are considered fundamental units of brain activity, underlying diverse functions ranging from perception to memory and decision-making. Cell assemblies have historically been conceived of as internal representations of specific stimuli or actions. Alternatively, cell assemblies can be endogenously defined by their ability to effectively elicit specific responses in downstream ('reader') neurons. Yet, whether cell assemblies are selectively detected by downstream neurons remains unknown. Here, we provide evidence for such assembly-reader communication. Reader activation was genuinely collective, functionally selective, yet flexible, implementing both pattern separation and completion. These processes occurred at the time scale of membrane integration, synaptic plasticity and gamma oscillations. Finally, assembly-reader couplings were selectively modified upon learning, indicating that they were plastic and reflected behaviorally relevant variables. These results support cell assemblies as an endogenous mechanism for brain function.

The ability to efficiently find alternatives routes when faced with unexpected obstacles along our path is among the most compelling evidence of the flexibility of human behaviour. Although a plethora of plausible computations have been put forward to elucidate how the brain accomplishes efficient goal-oriented navigation, the mechanisms that guide an effective re-planning when facing obstructions are still largely undetermined. There is a fair consensus in postulating that possible alternatives routes are internally replayed sampling from past experiences, however, there is currently no account of the criterion according to which those memories are replayed. Here, we posit that paths, which are expected to be more rewarding are replayed more often and that eye movements are the explicit manifestation of this re-planning strategy. In other words, the visual sampling statistics reflects the retrieval of available routes on a mental representation of the environment. To test our hypothesis, we measured the ability of both young and old human subjects to solve a virtual version of the Tolman maze, while we recorded their eye movements. We used reinforcement learning (RL) to corroborate that eye movements statistics was crucially subtending the decision making process involved in re-planning and that the incorporation of this additional information to the algorithm was necessary to reproduce the behavioral performance of both screened populations.