Chunking as a rational strategy for lossy data compression in visual working memory

As part of the Past Neuroeconomics Colloquium Spring 2017

Speaker: Prof. Michael Frank, Ph.D.

Professor of Cognitive, Linguistic & Psychological Sciences,
Brown University

Chunking as a rational strategy for lossy data compression in visual working memory

Time: 2:40 pm to 4:00 pm
Date: Tuesday, April 25th, 2017
Location: NYU Department of Economics, 19 W 4th Street, Room 517




The amount of visual information that can be stored in working memory is inherently limited, and the nature of this limitation has been a subject of intense debate. The debate has relied on tasks and models that assume visual items are independently encoded in working memory. Here we propose an alternative to this assumption: similar features are jointly encoded through a ‘chunking’ process to optimize performance on visual working memory tasks. We show that such chunking can: 1) facilitate performance improvements for abstract capacity-limited systems, 2) be optimized through reinforcement learning, 3) be implemented by neural network center-surround dynamics, and 4) increase effective storage capacity at the expense of recall precision. Human subjects performing a delayed report working memory task show evidence of the performance advantages, trial-to-trial behavioral adjustments, precision detriments, and inter-item dependencies predicted by optimization of task performance though chunking. Furthermore, by applying similar analyses to previously published datasets, we show that markers of chunking behavior are robust and increase with memory load. Taken together, our results support a more nuanced view of visual working memory capacity limitations: tradeoff between memory precision and memory quantity through chunking leads to capacity limitations that include both discrete (item limit) and continuous (precision limit) aspects.

Full Paper: Click here

About the Speaker

Michael J. Frank, PhD is Professor of Cognitive, Linguistic & Psychological Sciences and
Psychiatry and Human Behavior and is affiliated with the Brown Institute for Brain Science at
Brown University. He directs the Brown Initiative for Computation in Brain and Mind and his own Laboratory for Neural Computation and Cognition. He received his PhD in Neuroscience and Psychology in 2004 at the University of Colorado, following undergraduate and master’s degrees in electrical engineering and biomedicine (Queen’s University (Canada) and University of Colorado). Dr. Frank’s work focuses primarily on theoretical models of frontostriatal circuits and their modulation by dopamine, especially in terms of their cognitive functions and implications for neurological and psychiatric disorders. The models are tested and refined with experiments involving pharmacological manipulation, deep brain stimulation, EEG, fMRI and genetics. Awards include the Cognitive Neuroscience Society Young Investigator Award (2011), the Janet T Spence Award for early career transformative contributions (Association for Psychological Science, 2010) and the DG Marquis award for best paper published in Behavioral Neuroscience (2006). Dr Frank is a Kavli Fellow (2016), a member of Faculty of 1000 (Theoretical Neuroscience section), and serves as an editor for eLife, the Journal of Neuroscience, and Behavioral Neuroscience.