A connectionist architecture of visuospatial working memory

16:30 - 18:00

Mon-B22-Talk III-

Computational Approaches to Modeling Cognition

Chair/s:

Thorsten Pachur, Chris Donkin

Computational modeling provides a powerful tool to study and measure the cognitive underpinnings of behavior. This symposium features recent advances in the application of computational modeling in experimental psychology, showcasing its immense value for learning about cognitive processing across a wide range of applications. Florian Bolenz presents an analysis with the computational framework of metareasoning to model differences between younger and older adults in boundedly rational strategy selection during risky choice. The contribution by Ann-Katrin Hosch features a new evidence-accumulation exemplar model of category learning that allows an examination of how the variance of sampled examplars influences categorization. Chris Donkin presents a project that uses computational modeling to distinguish basic memory processes and strategic response in the DRM paradigm, highlighting the often neglected role of reasoning processes in recognition memory research. Veronika Zilker integrates attentional processes in the computational modeling of decision making with cumulative prospect theory; specifically, she examines whether attentional processes might be key drivers of the description-experience gap in risky choice. In the contribution by David Izydorczyk, a blending model of exemplar-based and rule-based judgment is used to model the cognitive processes underlying quantitative judgment of complex stimuli. Benjamin Kowialiewski presents a connectionist model of visuospatial working memory to study the impact of visuospatial proximity on memory performance. The symposium will bring together researchers from various research groups in Europe, reflecting the increasing popularity of cognitive modeling in experimental psychology.

Back

A connectionist architecture of visuospatial working memory

Mon-B22-Talk III-06

Presented by: Benjamin Kowialiewski

Benjamin Kowialiewski ¹, Steve Majerus ², Klaus Oberauer ¹

¹ University of Zurich, ² University of Liège

We present a connectionist model of visuospatial working memory (WM). The core WM architecture encodes new information by binding it to contexts through Hebbian learning. The representations encoded by the model are 2-dimensional spatial locations. These representations were created using the internal pattern of activation from an auto-encoder that learns to reproduce its input. We simulated an experiment in which the model must encode locations of varying proximity presented sequentially, followed by order reconstruction and recall tests. The model generates two important predictions. First, spatial proximity impairs memory for order: In an order reconstruction test, WM representation of spatially closer locations are more difficult to discriminate, leading to increased confusion errors. Second, spatial proximity improves memory for items: In a recall task, the recall error (Euclidean distance) is smaller in sequences composed of spatially close locations. We tested the model's predictions against data from 30 subjects who were asked to perform the same task as the model. The two predictions from the model were confirmed. We propose that similarity effects in WM are governed by domain-general principles, as equivalent observations have been established for other dimensions of similarity, such as the auditory, visual, and phonological similarity.

Keywords: Modeling, Connectionism, Working memory, Visuospatial

Bookmark

Back