Neural signature of flexible coding in prefrontal cortex

Abstract

The ability of prefrontal cortex to quickly encode novel associations is crucial for adaptive behavior and central to working memory. Fast Hebbian changes in synaptic strength permit forming new associations, but neuronal signatures of this have been elusive. We devised a trialwise index of pattern similarity to look for rapid changes in population codes. Based on a computational model of working memory, we hypothesized that synaptic strength—and consequently, the tuning of neurons—could change if features of a subsequent stimulus need to be “reassociated,” i.e., if bindings between features need to be broken to encode the new item. As a result, identical stimuli might elicit different neural responses. As predicted, neural response similarity dropped following rebinding, but only in prefrontal cortex. The history-dependent changes were expressed on top of traditional, fixed selectivity and were not explainable by carryover of previous firing into the current trial or by neural adaptation.

Domain general frontoparietal regions show modality-dependent coding of auditory and visual rules

“…In these models, general-purpose “conjunction” units come to represent conjunctions of stimuli due to fixed random recurrent connections (Buschman, 2021) or flexibly through rapid Hebbian updating of synaptic weights that form bespoke conjunctions depending on the current task (Manohar et al, 2019). Accordingly, neurons in these regions have been shown to respond to a conjunction of different sensory inputs, in different contexts and at particular points in time (Mante, Sussillo, Shenoy, & Newsome, 2013; Rigotti et al, 2013; Aoi, Mante, & Pillow, 2020; Bocincova et al, 2022). These mixed-selectivity properties result in high dimensional spaces for representing cognitive variables and maintaining a unique combination of inputs, capturing the diversity of information from different brain regions (Bouchacourt & Buschman, 2019; Badre et al, 2021; Buschman, 2021).…”

“…Third, domain generality could be defined as when a network re-uses the same resources to process these different types of information in different situations. This could arise if general-purpose resources are flexibly allocated to process information from each of the two modality-specific tasks (for example, as in Manohar, Zokaei, Fallon, Vogels, & Husain, 2019; Bocincova, Buschman, Stokes, & Manohar, 2022). This is challenging to assess with fMRI (as even our highest resolution protocols capture the activity of tens of thousands of neurons) but we can at least ask whether patterns from multiple modalities load onto the same or different voxels (Jackson & Woolgar, 2018).…”

Changes in behavioral priority influence the accessibility but not the quality of working memory content

“…Likewise, in the case of switching between multiple potentially relevant WM items, priority may refer to the (internal) selection of likely task-relevant WM content. Research emphasizing transformations in the neural representations of WM content following changes in behavioral relevance (e.g., Panichello & Buschman, 2021; Bocincova et al, 2022; Li & Curtis, 2022) may help extirpate the use of colloquial terms like “selection”, “retrieval”, and “priority”.…”

Prefrontal cortex temporally multiplexes slow and fast dynamics in value learning and memory

“…Figure 5A shows the time-frequency power modulations for good and bad objects and their difference (LFP value signal, bottom row) in PV task in the four time points and in different frequency bands. Almost all major band powers including high-Gamma (60-200 Hz), Beta (12-30 Hz), Alpha (8)(9)(10)(11)(12) and Theta (4-7 Hz) showed significant modulation in response to objects and many showed differences for good vs bad object presentations. Among all frequencies the power in high-Gamma resembled the pattern seen in the population average the most with temporal multiplexing of early and late components of its value signal (Fig.…”

“…Among various cortical areas, the ventrolateral prefrontal cortex (vlPFC), is a potential candidate to handle the balance between stability vs flexibility in value learning. vlPFC is long recognized for its role in encoding short-term value associated with objects [10][11][12][13] , but more recently its involvement in representing long-term memory of object values is uncovered 14 . In addition to its multifaceted function, vlPFC is among few areas that take part in both cortical-striatal-thalamocortical loops that control flexibility and stability in reward learning and memory 15,16 .…”