Hierarchical representation of multi-step tasks in multiple-demand and default mode networks

Abstract

† Joint senior authors Number of pages: 37 Number of figures: 6 (Images of faces are covered to comply with bioRxiv requirements) Number of words for abstract: 230 Number of words for introduction: 649 Number of words for discussion: 1467The authors declare no competing financial interests. Abstract 27Task episodes consist of sequences of steps that are performed to achieve a goal. We used fMRI to 28 examine neural representation of task identity, component items, and sequential position, focusing on two 29 100 stew"), steps (e.g. third) and items associated with steps (e.g. "wash vegetables"). We used univariate 101 event-related and finite impulse response (FIR) analyses to characterize the temporal evolution of activity 102 across episodes, and representational similarity analysis (RSA) to investigate representations of task 103 structure and content. We hypothesized that different brain regions would be preferentially sensitive to 104 different levels of the temporal task hierarchy, and focused on the MD and DMN networks as a priori 105 regions of interest. 106 107 Task episode representation in cortical networks 6 Methods 108 Participants 109 42 participants (20 male, 22 female; ages 18-39, mean = 26.79, SD = 4.77) were included in the 110 experiment at the MRC Cognition and Brain Sciences Unit. An additional 19 participants were excluded 111(two were discovered to have cysts, one lost several slices due to poor bounding box positioning, ten were 112 excluded due to having no correct episodes for at least one combination of cued task × distractor task (see 113 later), and a further six were excluded due to excessive head motion > 5 mm). All participants were 114 neurologically healthy, right-handed, with normal or corrected-to-normal vision. Procedures were carried 115 out in accordance with ethical approval obtained from the Cambridge Psychology Research Ethics 116 Committee, and participants provided written, informed consent before the start of the experiment 117 118Stimuli and task procedures 119 The study consisted of a learning session outside the scanner and an execution session in the scanner. 120During the learning session, participants learned six everyday task sequences, each based in one of two 121 locations ("rooms"; three kitchen and three bathroom). Each task consisted of four ordered "steps". For 122 example, the task "make a stew" consisted of the steps "take food from fridge", "wash vegetables", "chop 123 vegetables", "cook on stove". Each step was associated with a unique image ("item"). The complete set 124 of stimuli is shown in Figure 1A. 125(e.g., "make a stew"). After a short delay, the first search array of four items appeared, and participants 132 were asked to select the item corresponding to the first step of that task (here, "take food from fridge"). 133 Participants selected this same target in three search arrays (total step duration = 9 s), then were given 134 a brief indicator that the step had been completed, and moved on to the next step (here "wash 135 vegetables"). ...

Preferential activation of the posterior Default-Mode Network with sequentially predictable task switches

“…More specifically, those two studies demonstrated a fractionation in DMN when large perceptual switches occurred in the task, with the activation of the PCC/PrCC, PHC, vmPFC and amPFC co-occurring with that sub-regions of the multiple-demand network. Network re-configuration has also been observed in IBL, where presentation of new rules results in global signal elevation across distributed brain regions including the DMN [27,55], and during a multi-step task with the initiation of a new episode [66]. This may be indicative of an increase in resource-allocation in response to increase cognitive demands.…”

“…Notably, in both cases, DMN sub-regions coactivate alongside task positive networks that they characteristically anti-correlate with [11,22,23]. More recent evidence has suggested that the DMN and the multiple demand cortex (MDC), which spans frontal and parietal regions that commonly are activated during challenging cognitive tasks, serve complementary roles in tasks where multi-step decision making takes place [66]. How this fractionation of the DMN and its interaction with task-positive networks relates to the distinct executive sub-processes of switching, remains to be determined.…”

The default mode network in cognition: a topographical perspective

“…Other studies have found that activity in DMN regions is related to relatively abstract features of cognition, such as the level of specificity 71,[89][90][91] or vividness 92,93 of a stimulus. Finally, studies have also shown that during external, goal-orientated thought, the DMN represents features of task context, rather than the specific details of the steps needed to achieve a goal 94,95 . Together, these studies are consistent with the view that, whereas concrete features of cognition may depend on peripheral brain systems, neural activity within regions of the DMN may reflect more abstract features of cognition.…”

Comprehension of computer code relies primarily on domain-general executive brain regions

“…Furthermore, given that no regions outside of the MD system showed code-specific responses, it must be the case that code-specific knowledge representations are also stored within this system (see Hasson et al, 2015, for a general discussion of the lack of distinction between storage and computing resources in the brain). Much evidence suggests that the MD system can flexibly store task-relevant information in the short term (e.g., Fedorenko et al, 2013;Freedman et al, 2001;Shashidhara, Mitchell, et al, 2019;Wen et al, 2019;Woolgar et al, 2011). However, evidence from studies on processing mathematics (e.g., Amalric & Dehaene, 2019) and physics (e.g., Cetron et al, 2019;Fischer et al, 2016) further suggests that the MD system can store some domain-specific representations in the long term, perhaps for evolutionarily lateemerging and ontogenetically late-acquired domains of knowledge.…”