The problem
After several rounds sorting cards by color, switch to sorting by shape. A red circle now belongs with blue circles instead of red triangles. You must keep the new rule in mind and let it change which visible feature guides your response. How do interacting brain systems make that change while other information remains present?[1]
Cognitive control lets a goal change which information guides a decision and which action follows. How do interacting brain circuits maintain that goal and change their processing when the rule changes? One experiment gives a concrete case: macaques saw moving, colored dots, and a cue told them to judge either motion direction or color. The same display called for different decisions depending on the cue.
In Mante and colleagues' model, connections let activity feed back through the network; this is a recurrent circuit. Changing the context altered how the network's collective activity evolved. A brief relevant input had a lasting effect on the model's developing decision, while the effect of an irrelevant input was largely reversed by the network's dynamics.
This supplies a proposed mechanism for selecting evidence within a circuit that receives both features. The model reproduced important aspects of the monkey recordings, but its trained connections are not a measurement of the animals' complete wiring. The result is scoped to the studied color–motion task and frontal population.[1] [2]
Scope
The interacting neural mechanisms that maintain goals and adjust processing or actions when tasks and circumstances change.
What remains open
In one experiment, monkeys used a cue to judge either color or motion in a display containing both. Recordings showed that information about the irrelevant feature could still reach the studied frontal circuit. The problem includes how a rule changes which information influences a decision, and how that process fits with maintaining goals and switching between them in broader behavior.[1] [2]
Human network studies show groups of brain regions whose activity changes together. This functional connectivity identifies relationships among signals; it does not by itself show a direct connection or how one region causes another to change. Assigning a control function to a network therefore leaves the mechanism incomplete.
Open questions include how prefrontal networks interact with deeper structures such as the basal ganglia and thalamus, how those interactions unfold during a task, and which computations are shared across maintaining, updating and stopping responses. Circuit interventions, recordings and models must connect these levels without treating one laboratory task as a complete account of cognitive control.[1] [3]
Why it matters
Holding to a rule lets you carry out a sequence reliably; updating it lets you use the same skill when the situation changes. These needs can pull in different directions. Understanding their coordination would connect flexible behavior with everyday errors in which an old response persists after the goal has changed.[1]
A goal must influence processing long enough to organize several actions, yet allow new evidence or instructions to change what happens next. A mechanism that handles both demands would explain how cooperating circuits can support stable plans and flexible responses using overlapping resources. This would connect distributed brain coordination with flexible behavior, without assuming a single central controller. It also gives specific models a broader test: do they explain both persistence and appropriate change across tasks?[1] [2]
Definitions
2 terms
recurrent circuit
A circuit with connections through which activity feeds back into the network. In the model of the color–motion task, recurrent dynamics help select relevant evidence; those trained connections are a proposed mechanism, not a recovered map of the monkey circuit.[2]
functional connectivity
A statistical relationship between activity signals from different brain regions, often measured by how they change together. It does not alone establish a direct anatomical connection or a causal influence between those regions.[1]
Sources
- The role of PFC networks in cognitive control and executive function
Notes on The role of PFC networks in cognitive control and executive function
Research review of interacting prefrontal and distributed networks, functional connectivity, temporal dynamics and open cortical–subcortical mechanisms. Published online 18 August 2021; the January 2022 issue is separate. Clinical proposals are not adopted.
- Context-dependent computation by recurrent dynamics in prefrontal cortex
Notes on Context-dependent computation by recurrent dynamics in prefrontal cortex
Two macaques performed a cued color–motion decision task; population recordings and a trained recurrent model support a proposed mechanism for selecting evidence. Model perturbations are not interventions on the animals or identification of their full connectivity.
- Selective gating of visual signals by microstimulation of frontal cortex
Notes on Selective gating of visual signals by microstimulation of frontal cortex
Primary stimulation experiment linking frontal eye field activity with spatially corresponding visual-area V4 responses below the threshold for an eye movement. This is a scoped circuit effect, not a complete mechanism of goal maintenance or rule switching.