Renata Cruz^1,2 (), Thomas Christophel^1,2; ¹Humboldt-Universität zu Berlin, Berlin, Germany, ²Bernstein Center for Computational Neuroscience and Berlin Center for Advanced Neuroimaging, Berlin, Germany

The ability to actively remove irrelevant items from working memory storage is essential for goal-directed behaviour. However, the mechanisms underlying active removal of information remain largely unknown. Importantly, exerting control over attentional selection of a memorized stimulus requires information about this specific selection: Which stimulus should be memorized, and which stimulus should be forgotten? Our recent work provides evidence that this information is represented in the form of selection-specific activity in parietal and temporal regions, suggesting that control processes might be distributed across different cortical regions. Here, we employ fMRI and multivariate pattern analyses (using cross-validated (Cross-)MANOVA) to identify selection-specific across the cortex through a continuous task. In this modified version of an n-Back discrimination task, subjects were tasked with memorizing an item until a coloured retro-cue indicated if that item should be remembered, and used for an upcoming task, or forgotten. Thus, while prior work investigated which process underly the selection between two items, the current study focusses on the fate of a singular item. Sample stimuli could be auditory complex tones or visual gratings. To identify areas that carry the information conveyed by the cue as to whether an item should be actively removed from working memory, we performed a searchlight analysis and estimated variance explained by the cue (‘remember’ versus ‘forget’) across the brain. We find a wide array of frontal and parietal brain regions for holding this information including precentral and lateral prefrontal cortex as well posterior and medial parietal regions. These results highlight that the active removal of working memory contents is unlikely to rely on one individual patch of cortex. Cross-regional interactions need to be understood to distinguish the unique contribution of these individual circuits and modality-specific control processes need to be distinguished for general ones, if those exist.