A Diversity of Intrinsic Timescales Underlie Neural Computations

Abstract

Neural processing occurs across a range of temporal scales. To facilitate this, the brain uses fast-changing representations reflecting momentary sensory input alongside more temporally extended representations, which integrate across both short and long temporal windows. The temporal flexibility of these representations allows animals to behave adaptively. Short temporal windows facilitate adaptive responding in dynamic environments, while longer temporal windows promote the gradual integration of information across time. In the cognitive and motor domains, the brain sets overarching goals to be achieved within a long temporal window, which must be broken down into sequences of actions and precise movement control processed across much shorter temporal windows. Previous human neuroimaging studies and large-scale artificial network models have ascribed different processing timescales to different cortical regions, linking this to each region’s position in an anatomical hierarchy determined by patterns of inter-regional connectivity. However, even within cortical regions, there is variability in responses when studied with single-neuron electrophysiology. Here, we review a series of recent electrophysiology experiments that demonstrate the heterogeneity of temporal receptive fields at the level of single neurons within a cortical region. This heterogeneity appears functionally relevant for the computations that neurons perform during decision-making and working memory. We consider anatomical and biophysical mechanisms that may give rise to a heterogeneity of timescales, including recurrent connectivity, cortical layer distribution, and neurotransmitter receptor expression. Finally, we reflect on the computational relevance of each brain region possessing a heterogeneity of neuronal timescales. We argue that this architecture is of particular importance for sensory, motor, and cognitive computations.

Central Thalamic-Medial Prefrontal Control of Adaptive Responding in the Rat: Many Players in the Chamber

“…These findings are consistent with the view that different regions of PFC support different domains of working memory (Goldman-Rakic, 2005;Kesner and Churchwell, 2011). More recent evidence has shown that different modalities of working memory are mediated by other areas of the association cortex (Miller et al, 2018;Xu, 2018;Buchsbaum and D'Esposito, 2019;Cavanaugh et al, 2020) indicating that working memory may represent a property of cortex that extends beyond PFC.…”

“…Working memory is important for adaptive responding: allowing organisms to hold and manipulate information online while monitoring fluctuations in the environment (Miller et al, 2018;Cavanaugh et al, 2020). Early lesion studies indicated that mPFC lesions in the rat have a selective effect on working memory, impairing delayed conditional discriminations based on response-related egocentric information while sparing maze tasks that require an allocentric solution (Becker et al, 1980;Kolb et al, 1983;Kesner et al, 1996;de Bruin et al, 2001).…”

Memory and Markov Blankets

“…While the duration of a memory appears to be on a continuum, based upon the analysis above, there are biological considerations that favor the separation of time scales found in psychology. Specifically, populations of neurons have time constants, in the order of microseconds to hundreds of milliseconds [ 43 ], that are much shorter than the time constants associated with changes in synaptic efficacy, in the order of minutes to hours [ 44 ]—although there is considerable variation within each of these categories. Note, for example, the relatively small fluctuations in the average trajectory plot in Figure 1 , which is considerably smaller than the fluctuations in the trajectories over which this average was constructed.…”

“…This is defined as memory that persists over seconds in which a memory is held ‘online’ for the purposes of some cognitive processing (e.g., for planning) [ 46 ]. It is often associated with structures like the prefrontal cortex [ 8 , 43 , 47 ], where neural activity evolves over longer durations than in sensory cortical regions [ 48 ]. The inclusion of a behavioral aspect in definitions of working memory hints at why it is useful to maintain direct conditional dependencies between internal and external states.…”

Frontal circuit specialisations for decision making

“…More recently, similar organising principles of resting time constants have also been found in the rodent (Siegle et al., 2021) and human brain (Demirtaş et al., 2019; Gao et al., 2020; Vidaurre et al., 2018). Several studies have also shown that neurons with longer resting time constants are more likely to be functionally involved in temporally extended cognitive tasks (Cavanagh et al., ,2016, 2018, 2020; Fascianelli et al., 2019; Nishida et al., 2014; Wasmuht et al., 2018).…”