NeuroTalk: Chunking of global and local representations
共 2546字,需浏览 6分钟
·
2021-09-27 00:49
本文来源:Neurotalk
Online Symposium
Evidence for chunking of global and local representations in working memory
Presenter: Mr. Yuhang Li
Objects stored in working memory (WM) in daily life contain hierarchical structures with global shape and local details. Previous research regarding hierarchical representations in WM mainly focused on how global and local targets are processed differently, which often found that global representations are always better remembered. However, it remains unknown whether global and local representations would interact in WM. In the present study, instead of focusing on the differences between global and local representations, we investigated whether and how global and local representations would be stored jointly in WM. Our evidence showed that hierarchical features were not stored independently. Differences between global and local orientations modulate their corresponding memory fidelity. To investigate the cognitive mechanisms underlying the interactions between hierarchical memory representations, we conducted factorial model comparison to two delay estimation datasets across 16 models with 4 factors: 1) variable working memory precision across trials, 2) bias towards different levels of the cued memory item, 3) swapping between levels of the cued memory item, 4) chunking between global and local representations. This modeling procedure revealed that models with chunking and variable precision best explained the data. These findings implied an automatic chunking but not bias or swapping process between different levels of hierarchical representations in WM.
Bio
Mr. Yuhang Li obtained his bachelor’s degree from the department of psychology in East China Normal University. Currently, he is a Phd student in Psychology at Prof. Qiyang Nie’s Computational Psychophysiology (CPP) lab at the Centre for Cognitive and Brain Sciences and Faculty of Social Sciences in the University of Macau. Yuhang uses computational modeling method to investigate the cognitive mechanism underlying a variety of working memory tasks and decision tasks.
Moderator: Ms. Christy, Jianhua Li
(PhD student, CCBS/FSS, University of Macau)
Time: 7:00 PM, Sep 24th
Zoom ID: 959 6270 3011
Passcode: ccbs
Acknowledgements
NeuroTalk x Brain Technology
Scan the QR code to sign up as Presenter!
Or contact Mr. Gao Fei (yb87711@um.edu.mo) for inquiries.