Embedding stochastic dynamics of the environment in spontaneous activity by prediction-based plasticity DOI Open Access
Toshitake Asabuki, Claudia Clopath

Published: Dec. 3, 2024

The brain learns an internal model of the environment through sensory experiences, which is essential for high-level cognitive processes. Recent studies show that spontaneous activity reflects such learned model. Although computational have proposed Hebbian plasticity can learn switching dynamics replayed activities, it still challenging to dynamic obeys statistical properties experience. Here, we propose a pair biologically plausible rules excitatory and inhibitory synapses in recurrent spiking neural network embed stochastic activity. synaptic rule seeks minimize discrepancy between stimulus-evoked internally predicted activity, while maintains excitatory-inhibitory balance. We reactivation cell assemblies follows transition statistics model’s evoked dynamics. also demonstrate simulations our replicate recent experimental results songbirds, suggesting might underlie mechanism by animals models environment.

Language: Английский

Embedding stochastic dynamics of the environment in spontaneous activity by prediction-based plasticity DOI Open Access
Toshitake Asabuki, Claudia Clopath

Published: April 12, 2024

The brain learns an internal model of the environment through sensory experiences, which is essential for high-level cognitive processes. Recent studies show that spontaneous activity reflects such learned model. Although computational have proposed Hebbian plasticity can learn switching dynamics replayed activities, it still challenging to dynamic obeys statistical properties experience. Here, we propose a pair biologically plausible rules excitatory and inhibitory synapses in recurrent spiking neural network embed stochastic activity. synaptic rule seeks minimize discrepancy between stimulus-evoked internally predicted activity, while maintains excitatory-inhibitory balance. We reactivation cell assemblies follows transition statistics model’s evoked dynamics. also demonstrate simulations our replicate recent experimental results songbirds, suggesting might underlie mechanism by animals models environment.While often seen as simple background noise, work has hypothesized instead brain’s learnt While several generate structured learning embedding unclear. Using model, investigate obeying appropriate statistics. Our shed light on crucial step towards better understanding role generative processes complex environments.

Language: Английский

Citations

5
Embedding stochastic dynamics of the environment in spontaneous activity by prediction-based plasticity DOI Creative Commons
Toshitake Asabuki, Claudia Clopath

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: May 1, 2023

Abstract The brain learns an internal model of the environment through sensory experiences, which is essential for high-level cognitive processes. Recent studies show that spontaneous activity reflects such learned model. Although computational have proposed Hebbian plasticity can learn switching dynamics replayed activities, it still challenging to dynamic obeys statistical properties experience. Here, we propose a pair biologically plausible rules excitatory and inhibitory synapses in recurrent spiking neural network embed stochastic activity. synaptic rule seeks minimize discrepancy between stimulus-evoked internally predicted activity, while maintains excitatory-inhibitory balance. We reactivation cell assemblies follows transition statistics model’s evoked dynamics. also demonstrate simulations our replicate recent experimental results songbirds, suggesting might underlie mechanism by animals models environment.

Language: Английский

Citations

2
Learning predictive signals within a local recurrent circuit DOI Creative Commons
Toshitake Asabuki, Colleen J. Gillon, Claudia Clopath

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: June 15, 2023

Abstract The predictive coding hypothesis proposes that top-down predictions are compared with incoming bottom-up sensory information, prediction errors signaling the discrepancies between these inputs. While this explains presence of errors, recent experimental studies suggest error signals can emerge within a local circuit, is, from input alone. In paper, we test whether circuits alone generate by training recurrent spiking network using plasticity rules. Our model replicates resembling various results, such as biphasic pattern and context-specific representation signals. findings shed light on how synaptic shape enables acquisition updating an internal neural network.

Language: Английский

Citations

1
Embedding stochastic dynamics of the environment in spontaneous activity by prediction-based plasticity DOI Open Access
Toshitake Asabuki, Claudia Clopath

Published: April 12, 2024

The brain learns an internal model of the environment through sensory experiences, which is essential for high-level cognitive processes. Recent studies show that spontaneous activity reflects such learned model. Although computational have proposed Hebbian plasticity can learn switching dynamics replayed activities, it still challenging to dynamic obeys statistical properties experience. Here, we propose a pair biologically plausible rules excitatory and inhibitory synapses in recurrent spiking neural network embed stochastic activity. synaptic rule seeks minimize discrepancy between stimulus-evoked internally predicted activity, while maintains excitatory-inhibitory balance. We reactivation cell assemblies follows transition statistics model’s evoked dynamics. also demonstrate simulations our replicate recent experimental results songbirds, suggesting might underlie mechanism by animals models environment.While often seen as simple background noise, work has hypothesized instead brain’s learnt While several generate structured learning embedding unclear. Using model, investigate obeying appropriate statistics. Our shed light on crucial step towards better understanding role generative processes complex environments.

Language: Английский

Citations

0
Embedding stochastic dynamics of the environment in spontaneous activity by prediction-based plasticity DOI Open Access
Toshitake Asabuki, Claudia Clopath

Published: Dec. 3, 2024

The brain learns an internal model of the environment through sensory experiences, which is essential for high-level cognitive processes. Recent studies show that spontaneous activity reflects such learned model. Although computational have proposed Hebbian plasticity can learn switching dynamics replayed activities, it still challenging to dynamic obeys statistical properties experience. Here, we propose a pair biologically plausible rules excitatory and inhibitory synapses in recurrent spiking neural network embed stochastic activity. synaptic rule seeks minimize discrepancy between stimulus-evoked internally predicted activity, while maintains excitatory-inhibitory balance. We reactivation cell assemblies follows transition statistics model’s evoked dynamics. also demonstrate simulations our replicate recent experimental results songbirds, suggesting might underlie mechanism by animals models environment.

Language: Английский

Citations

0