Attractor dynamics gate cortical information flow during decision-making

Abstract

One Sentence Summary: Mechanisms controlling state-dependent communication between brain regions allow for robust action-selection. AbstractDecisions about future actions are held in memory until enacted, making them vulnerable to distractors. The neural mechanisms controlling decision robustness to distractors remain unknown.We trained mice to report optogenetic stimulation of somatosensory cortex, with a delay separating sensation and action. Distracting stimuli influenced behavior less when delivered later during delay -demonstrating temporal gating of sensory information flow. Gating occurred even though distractor-evoked activity percolated through the cortex without attenuation. Instead, choicerelated dynamics in frontal cortex became progressively robust to distractors as time passed.Reverse-engineering of neural networks trained to reproduce frontal-cortex activity revealed that chosen actions were stabilized via attractor dynamics, which gated out distracting stimuli. Our results reveal a dynamic gating mechanism that operates by controlling the degree of commitment to a chosen course of action.

Rotational dynamics in motor cortex are consistent with a feedback controller

“…The hand area of MC also does not exhibit rotational dynamics during grasping-only behaviour (Suresh et al, 2020), though it does exhibit rotational dynamics during reach-to-grasp (Abbaspourazad et al, 2021;Vaidya et al, 2015) which may reflect the reaching component of the behaviour. More broadly there is a growing body of work characterizing cortical neural dynamics across different behavioural tasks which have revealed rotational (Abbaspourazad et al, 2021;Aoi et al, 2020;Gao et al, 2016;Kao et al, 2015;Libby and Buschman, 2021;Remington et al, 2018;Sani et al, 2021;Sohn et al, 2019;Stavisky et al, 2019;Vaidya et al, 2015), helical (Russo et al, 2020), stationary (Machens et al, 2010), and ramping dynamics (Finkelstein et al, 2021;Kaufman et al, 2016;Machens et al, 2010) and these dynamics appear to support various classes of computations. Thus, finding of rotational dynamics across the fronto-parietal circuit in the present study was not trivial.…”

A midbrain-thalamus-cortex circuit reorganizes cortical dynamics to initiate movement

“…Activity modes can be obtained by projecting neural activity along specific directions in neural state space, or similar dimensionality reduction methods (Cunningham and Yu, 2014;Kaufman et al, 2014;Kobak et al, 2016;Li et al, 2016). A successful decomposition of neural activity provides activity modes that are interpretable, by predicting specific aspects of behavior and revealing related neural computations (Mante et al, 2013;Kobak et al, 2016;Li et al, 2016;Inagaki et al, 2019;Finkelstein et al, 2019;Vyas et al, 2020;Lee and Sabatini, 2021).…”

Dynamical Feedback Control: Motor Cortex as an Optimal Feedback Controller Based on Neural Dynamics

“…In this view, the neural landscape must somehow include all likely movements, perturbations, and contexts: many narrow, parallel tracks separated by substantial walls and perhaps several shallow tracks all within one broad track. Perturbations push the neural state from one track to another more easily for those tracks with lower walls (Finkelstein et al, 2021). We propose that this is possible because of the ultra high-dimensional nature of motor cortex, with theoretically as many dimensions as there are neurons.…”