Latent representations in hippocampal network model co-evolve with behavioral exploration of task structure.

Latent representations in hippocampal network model co-evolve with behavioral exploration of task structure.
复制标题

海马网络模型中的潜在表示与任务结构的行为探索共同发展。

DOI:
10.1038/s41467-024-44871-6
复制
发表时间:
2024-01-23
影响因子:
16.6
通讯作者:
Clopath, Claudia
Clopath, Claudia
中科院分区:
综合性期刊1区
文献类型:
--
作者:
Cone, Ian;Clopath, Claudia

文献摘要

参考文献

相似文献

为了成功地学习现实生活中的行为任务,动物必须将行动或决定与任务的复杂结构配对,这可能取决于感官刺激和内部逻辑的抽象组合。众所周知,海马体会发展出这种复杂结构的表征,形成所谓的“认知地图”。然而,在种群水平上推动任务相关地图出现的确切生物物理机制仍不清楚。我们提出了一个模型,在该模型中,基于平台的学习在单细胞水平上,结合强化学习的代理,导致潜在的表征结构与行为的共同依赖地演变在特定任务的方式。与最近的实验数据相一致,我们表明,该模型成功地开发了潜在的结构必不可少的任务解决(线索依赖的“分裂器”),同时排除不相关的。最后,我们的模型进行了可测试的预测分裂表示和分裂行为政策之间的相互依赖的相互作用,在其演变过程中。单细胞可塑性机制如何导致任务依赖性认知地图仍不清楚。在这里,作者表明,海马体的这个模型表明,局部可塑性和行为的强化学习之间的合作可以导致特定于任务的潜在表征。
To successfully learn real-life behavioral tasks, animals must pair actions or decisions to the task’s complex structure, which can depend on abstract combinations of sensory stimuli and internal logic. The hippocampus is known to develop representations of this complex structure, forming a so-called “cognitive map”. However, the precise biophysical mechanisms driving the emergence of task-relevant maps at the population level remain unclear. We propose a model in which plateau-based learning at the single cell level, combined with reinforcement learning in an agent, leads to latent representational structures codependently evolving with behavior in a task-specific manner. In agreement with recent experimental data, we show that the model successfully develops latent structures essential for task-solving (cue-dependent “splitters”) while excluding irrelevant ones. Finally, our model makes testable predictions concerning the co-dependent interactions between split representations and split behavioral policy during their evolution. How mechanisms of single-cell plasticity lead to task-dependent cognitive maps remains unclear. Here, the authors show that this model of hippocampus shows that cooperation between local plasticity and reinforcement learning of behavior can lead to task-specific latent representations.
DOI: 10.7554/elife.73046
发表时间: 2021-12-09
期刊: eLife
影响因子: 7.7
作者:
Milstein AD;Li Y;Bittner KC;Grienberger C;Soltesz I;Magee JC;Romani S
通讯作者: Romani S
DOI: 10.1126/science.abm1670
发表时间: 2022-03-18
期刊: Science (New York, N.Y.)
影响因子: --
作者:
通讯作者: --
DOI: 10.1038/s41586-021-03652-7
发表时间: 2021-07
期刊: Nature
影响因子: 64.8
作者:
通讯作者: --
DOI: 10.1038/nature06725
发表时间: 2008-03-27
期刊: NATURE
影响因子: 64.8
作者:
Losonczy, Attila;Makara, Judit K.;Magee, Jeffrey C.
通讯作者: Magee, Jeffrey C.
DOI: 10.1016/0014-4886(76)90055-8
发表时间: 1976-01-01
影响因子: 5.3
作者:
OKEEFE, J
通讯作者: OKEEFE, J