Learning reward timing in cortex through reward dependent expression of synaptic plasticity

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• The neural basis of temporal representation in cognition remains largely unknown.
• Existing research often focuses on higher cortical areas, overlooking primary sensory cortices.
• A unifying theoretical framework for learning temporal representations is lacking.

Purpose of the Study:

• To propose a theoretical framework for how neural networks learn temporal representations.
• To investigate the role of synaptic plasticity in storing temporal information.
• To explain reward-time learning in the primary visual cortex (V1).

Main Methods:

• Development of a computational model based on reward-dependent synaptic plasticity.
• Numerical implementation of the model to simulate temporal learning.
• Analysis of lateral synaptic connections between neurons as storage sites for temporal representations.

Main Results:

• Demonstrated that local cortical networks can learn temporal representations through synaptic plasticity.
• Showed that reward-modulated plasticity is sufficient for learning these representations.
• Provided numerical evidence supporting the model's ability to explain reward-time learning in V1.

Conclusions:

• Temporal representations are stored in lateral synaptic connections.
• Reward-modulated synaptic plasticity is a key mechanism for learning temporal information.
• The proposed framework offers experimentally verifiable predictions for future research.