A tale of two lexica: Investigating computational pressures on word representation with neural networks

Frontiers in Artificial Intelligence

Artificial Intelligence

Enes Avcu¹  David W. Gow²  Kevin Scott Brown³  Michael Hwang⁴ 

[1] Department of Neurology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, United States;Department of Neurology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, United States;Athinoula A. Martinos Center for Biomedical Imaging Massachusetts General Hospital, Charlestown, MA, United States;Department of Psychology, Salem State University, Salem, MA, United States;Harvard-MIT Division of Health Sciences and Technology, Boston, MA, United States;Department of Pharmaceutical Sciences and School of Chemical, Biological, and Environmental Engineering, Oregon State University, Corvallis, OR, United States;Harvard College, Boston, MA, United States;

关键词: mental lexicon;    word representation;    neural networks;    functional segregation;    dorsal and ventral streams;    deep learning;   

DOI  :  10.3389/frai.2023.1062230

 received in 2022-10-05, accepted in 2023-03-10,  发布年份 2023

来源: Frontiers

【 摘 要 】

IntroductionThe notion of a single localized store of word representations has become increasingly less plausible as evidence has accumulated for the widely distributed neural representation of wordform grounded in motor, perceptual, and conceptual processes. Here, we attempt to combine machine learning methods and neurobiological frameworks to propose a computational model of brain systems potentially responsible for wordform representation. We tested the hypothesis that the functional specialization of word representation in the brain is driven partly by computational optimization. This hypothesis directly addresses the unique problem of mapping sound and articulation vs. mapping sound and meaning.ResultsWe found that artificial neural networks trained on the mapping between sound and articulation performed poorly in recognizing the mapping between sound and meaning and vice versa. Moreover, a network trained on both tasks simultaneously could not discover the features required for efficient mapping between sound and higher-level cognitive states compared to the other two models. Furthermore, these networks developed internal representations reflecting specialized task-optimized functions without explicit training.DiscussionTogether, these findings demonstrate that different task-directed representations lead to more focused responses and better performance of a machine or algorithm and, hypothetically, the brain. Thus, we imply that the functional specialization of word representation mirrors a computational optimization strategy given the nature of the tasks that the human brain faces.

【 授权许可】

Unknown   
Copyright © 2023 Avcu, Hwang, Brown and Gow.