Contributed Talk Sessions | Poster Sessions | All Posters | Search Papers

Poster Session B: Wednesday, August 13, 1:00 – 4:00 pm, de Brug & E‑Hall

Distributed Working Memory in a Computational Model of the Human Brain

Rares A. Dorcioman¹, Mengli Feng¹, Abhirup Bandyopadhyay², Jorge Mejias¹; ¹University of Amsterdam, ²Karolinska Institute Stockholm

Presenter: Rares A. Dorcioman

Working memory is a fundamental cognitive function that allows us to transiently store and manipulate relevant information in memory. While traditionally associated with localized prefrontal activity, recent electrophysiological and imaging studies reveal distributed activity across multiple brain regions. To uncover the mechanisms behind this distribution, we developed a detailed, data-constrained model of the human brain by integrating diverse large-scale datasets. Our model demonstrates that distributed working memory patterns emerge primarily through long-range synaptic projections rather than solely from local recurrent connectivity. We found that the network operates optimally near a critical region at the edge of a bifurcation, consistent with recent experimental and modeling findings, and explains approximately 60\% of the observed variability among brain areas involved in working memory. Furthermore, simulations of task-specific conditions, such as verbal and spatial working memory, indicate that high agreement with experimental data is achieved only when higher cortical regions modulate the network or when recurrent connectivity is enhanced across multiple circuits. These results suggest that working memory performance is the product of an interplay between distributed projections and context-dependent modulation, offering new insights into the neural substrates of human cognition.

Topic Area: Memory, Spatial Cognition & Skill Learning

Extended Abstract: Full Text PDF