Internal state dynamically gates task-specific attractor dynamics in prefrontal cortex

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• Internal states like motivation significantly impact cognitive functions, including working memory.
• Working memory relies on maintaining information over time and is known to be modulated by motivational factors.

Purpose of the Study:

• To investigate how motivational states influence the neural dynamics underlying working memory.
• To elucidate the mechanisms by which internal states modulate cognitive functions at the circuit level.

Main Methods:

• Combined population recordings of mouse medial prefrontal cortex (mPFC) neural activity with data-constrained recurrent neural network (RNN) modeling.
• Analyzed how task engagement affected attractor dynamics within a memory-maintenance subspace.

Main Results:

• Task engagement selectively modulated attractor dynamics supporting working memory, without altering stimulus-evoked responses.
• Recurrent neural network models revealed that engagement stabilized memory-specific attractors by reorganizing neural interactions.
• Continuous modulation of attractor geometry in both RNNs and mPFC predicted gradual changes in behavioral engagement.

Conclusions:

• Internal states modulate working memory by controlling the dynamical regime of medial prefrontal cortex circuits.
• This work provides a mechanistic link between internal states, neural dynamics, and cognitive function.
• Findings highlight the role of attractor dynamics in flexible cognitive control.