Probing the limits of activity-silent non-conscious working memory

Abstract

2 SummaryTwo types of working memory (WM) have recently been proposed: conscious active WM, depending on sustained neural activity, and activity-silent WM, requiring neither conscious awareness nor accompanying neural activity. However, whether both states support identical forms of information processing is unknown. Theory predicts that activity-silent states are confined to passive storage and cannot operate on stored information. To determine whether an explicit reactivation is required prior to the manipulation of information in WM, we evaluated whether participants could mentally rotate brief visual stimuli of variable subjective visibility. Behaviorally, even for unseen targets, subjects reported the rotated location above chance after several seconds. As predicted, however, such blindsight performance was accompanied by neural signatures of conscious reactivation at the time of mental rotation, including a sustained desynchronization in alpha/beta frequency and a decodable representation of participants' guess and response. Our findings challenge the concept of genuine non-conscious "working" memory, argue that activity-silent states merely support passive short-term memory, and provide a cautionary note for purely behavioral studies of non-conscious information processing.

Circuit mechanisms for the maintenance and manipulation of information in working memory

“…These studies suggest that tasks that require greater manipulation of the memoranda evoke greater levels of persistent neuronal activity, consistent with the correlation we observe between the level of manipulation and the level of persistent neuronal activity in our network models (Figure 7). These studies are also consistent with a recent human MEG study which also suggests that manipulating information in WM requires the reinstatement of persistent activity 42 .…”

Interplay between persistent activity and activity-silent dynamics in prefrontal cortex during working memory

“…However, our data also supports the idea that activity-silent and attractor-based mechanisms are not orthogonal, alternative mechanisms but they are largely co-expressed in the circuit and underlie different behaviors: whereas persistent attractor-based activity is engaged during active maintenance of memories, activity-silent maintenance supports secondary, possibly involuntary memory traces, here expressed in small serial biases. Similar ideas have been proposed in the context of attended and unattended memories 16,29,30,45 . Note, however, that in our proposed framework the close interplay between attractor-based and activity-silent mechanisms does not allow activity-silent memories to be protected from intervening attractor-based activations in the circuit.…”

A dopamine gradient controls access to distributed working memory in monkey cortex

“…By building a novel large-scale anatomically-constrained model of monkey cortex, we show how dopamine can engage robust distributed persistent activity mechanisms across connected higher cortical areas, and protect memories of behaviourally relevant-stimuli from distraction. As distributed persistent activity is necessary for the manipulation of thoughts in working memory (Masse et al 2019;Trübutschek et al 2019), dopamine release in the cortex may be a key step towards higher cognitive thought.…”

“…We show that short-term facilitation in long-range feedforward connections from early sensory areas to frontal and parietal cortex is a potential substrate for 'activity-silent' working memory in the absence of an initial prefrontal response to the stimulus. Given that experimental and modelling evidence suggests that manipulation of stored information requires a re-emergence of strong distributed activity (Masse et al 2019;Trübutschek et al 2019), the silent state may be better described as 'short-term memory', as noted by other authors (Mashour et al 2020;Masse et al 2020). The brain may then reserve widespread persistent activity for important information that must be used and manipulated to drive behavior.…”

“…This was possible regardless of whether the original stimulus resulted in prefrontal activity. Unconscious working memory results in similarly correct behavioural performance, without the characteristic frontoparietal delay-period activity (Trübutschek et al 2017(Trübutschek et al , 2019. Unconscious working memory is thought to rely on 'activity-silent' synaptic mechanisms (Stokes 2015;Trübutschek et al 2017).…”