Spontaneous emergence of alternating hippocampal theta sequences in a simple 2D adaptation model

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Июнь 10, 2024

SUMMARY Spatial sequences encoded by cells in the hippocampal-entorhinal region have been observed to spontaneously alternate across animal’s midline during navigation open field, but it is unknown how this occurs. We show that sinusoidal sampling patterns emerge rapidly and robustly a simple model of hippocampus makes no assumptions about sequence direction. corroborate our findings using hippocampal data from rats navigating field.

Язык: Английский

---

Sampling motion trajectories during hippocampal theta sequences

eLife, Год журнала: 2022, Номер 11

Опубликована: Ноя. 4, 2022

Efficient planning in complex environments requires that uncertainty associated with current inferences and possible consequences of forthcoming actions is represented. Representation has been established sensory systems during simple perceptual decision making tasks but it remains unclear if cognitive computations such as navigation are also supported by probabilistic neural representations. Here, we capitalized on gradually changing along planned motion trajectories hippocampal theta sequences to capture signatures representation population responses. In contrast prominent theories, found no evidence encoding parameters probability distributions the momentary activity recorded an open-field task rats. Instead, was encoded sequentially sampling randomly efficiently subsequent cycles from distribution potential trajectories. Our analysis first demonstrate hippocampus well equipped contribute optimal representing uncertainty.

Язык: Английский

---

Single spikes drive sequential propagation and routing of activity in a cortical network

eLife, Год журнала: 2023, Номер 12

Опубликована: Фев. 13, 2023

Single spikes can trigger repeatable firing sequences in cortical networks. The mechanisms that support reliable propagation of activity from such small events and their functional consequences remain unclear. By constraining a recurrent network model with experimental statistics turtle cortex, we generate temporally precise single spike triggers. We find rare strong connections sequence propagation, while dense weak modulate reliability. identify sections corresponding to divergent branches strongly connected neurons which be selectively gated. Applying external inputs specific the sparse backbone effectively control route within network. Finally, demonstrate concurrent interact reliably, generating highly combinatorial space activations. Our results reveal impact individual circuits, detailing how triggered, sustained, controlled during computations.

Язык: Английский

---

A recurrent network model of planning explains hippocampal replay and human behavior

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2023, Номер unknown

Опубликована: Янв. 19, 2023

Abstract When faced with a novel situation, humans often spend substantial periods of time contemplating possible futures. For such planning to be rational, the benefits behavior must compensate for spent thinking. Here we capture these features human by developing neural network model where itself is controlled prefrontal cortex. This consists meta-reinforcement learning agent augmented ability plan sampling imagined action sequences from its own policy, which call ‘rollouts’. The learns when beneficial, explaining empirical variability in thinking times. Additionally, patterns policy rollouts employed artificial closely resemble rodent hippocampal replays recently recorded during spatial navigation. Our work provides new theory how brain could implement through prefrontal-hippocampal interactions, are triggered – and adaptively affect dynamics.

Язык: Английский

---

Prioritizing replay when future goals are unknown

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Март 4, 2024

Abstract Although hippocampal place cells replay nonlocal trajectories, the computational function of these events remains controversial. One hypothesis, formalized in a prominent reinforcement learning account, holds that plans routes to current goals. However, recent puzzling data appear contradict this perspective by showing replayed destinations lag These results may support an alternative hypothesis updates route information build “cognitive map.” Yet no similar theory exists formalize view, and it is unclear how such map represented or what role plays computing it. We address gaps introducing learns candidate goals, before reward available when its location change. Our work extends planning account capture general map-building for replay, reconciling with data, revealing unexpected relationship between seemingly distinct hypotheses.

Язык: Английский

---