Home > Research > Publications & Outputs > Attention driven memory
View graph of relations

Attention driven memory

Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSNConference contribution/Paperpeer-review

Published

Standard

Attention driven memory. / Grunewalder, S.; Obermayer, K.
CogSci 2005: XXVII Annual Conference of the Cognitive Science Society July 21-23 Stresa, Italy. ed. / Bruno G. Bara; Lawrence Barsalou; Monica Bucciarelli. Cognitive Science Society, 2005. p. 845-850.

Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSNConference contribution/Paperpeer-review

Harvard

Grunewalder, S & Obermayer, K 2005, Attention driven memory. in BG Bara, L Barsalou & M Bucciarelli (eds), CogSci 2005: XXVII Annual Conference of the Cognitive Science Society July 21-23 Stresa, Italy. Cognitive Science Society, pp. 845-850. <http://csjarchive.cogsci.rpi.edu/Proceedings/2005/docs/p845.pdf>

APA

Grunewalder, S., & Obermayer, K. (2005). Attention driven memory. In B. G. Bara, L. Barsalou, & M. Bucciarelli (Eds.), CogSci 2005: XXVII Annual Conference of the Cognitive Science Society July 21-23 Stresa, Italy (pp. 845-850). Cognitive Science Society. http://csjarchive.cogsci.rpi.edu/Proceedings/2005/docs/p845.pdf

Vancouver

Grunewalder S, Obermayer K. Attention driven memory. In Bara BG, Barsalou L, Bucciarelli M, editors, CogSci 2005: XXVII Annual Conference of the Cognitive Science Society July 21-23 Stresa, Italy. Cognitive Science Society. 2005. p. 845-850

Author

Grunewalder, S. ; Obermayer, K. / Attention driven memory. CogSci 2005: XXVII Annual Conference of the Cognitive Science Society July 21-23 Stresa, Italy. editor / Bruno G. Bara ; Lawrence Barsalou ; Monica Bucciarelli. Cognitive Science Society, 2005. pp. 845-850

Bibtex

@inproceedings{817c3888310447a3bc56aa8a2750e846,
title = "Attention driven memory",
abstract = "Categorization is a skill which is used extensively in everyday life and as therefore an important aspect of human cognition. Consequently a variety of studies exist which address the topic and revealed that diverse factors affect human categorization performance. A critical but not extensively studied factor is time. Imagine watching a basketball game for 30 minutes. In this period of time plenty of actions will take place resulting in diverse impressions which make you afterwards categorize the game as interesting or boring. Such a categorization task is very similar to a time series classification task in the context of machine learning. In the field of machinelearning a phenomen called “vanishing gradient” is known which makes it generally hard to solve such a categorization task. A prominent method that overcomes this phenomen is the long short term memory which basicly consists of a memory that is controlled by two gate units which can be interpreted as adaptive encoding and recall units. Critical points which make the processing of the structure differ from human processing concern the encoding and the storage: (1) The structureis built to massively store information instead of carefully selecting few impressions for storage in memory.(2) Reweighting of stored information due to changing constellations is not possible. Coming back to the example this would mean that a nice action at the beginning of the game has a strong impact on your categorization independent of what kind of actions - might they be impressive or not - followed afterwards. In this work we tackle these points through introducing an attentionmechanism which drives the encoding and the storage of the structure. We analyse the model behavior in category learning tasks.",
keywords = "LSTM, Memory, Attention, Categorization, Modelling",
author = "S. Grunewalder and K. Obermayer",
year = "2005",
language = "English",
isbn = "0976831813",
pages = "845--850",
editor = "Bara, {Bruno G.} and Lawrence Barsalou and Monica Bucciarelli",
booktitle = "CogSci 2005",
publisher = "Cognitive Science Society",

}

RIS

TY - GEN

T1 - Attention driven memory

AU - Grunewalder, S.

AU - Obermayer, K.

PY - 2005

Y1 - 2005

N2 - Categorization is a skill which is used extensively in everyday life and as therefore an important aspect of human cognition. Consequently a variety of studies exist which address the topic and revealed that diverse factors affect human categorization performance. A critical but not extensively studied factor is time. Imagine watching a basketball game for 30 minutes. In this period of time plenty of actions will take place resulting in diverse impressions which make you afterwards categorize the game as interesting or boring. Such a categorization task is very similar to a time series classification task in the context of machine learning. In the field of machinelearning a phenomen called “vanishing gradient” is known which makes it generally hard to solve such a categorization task. A prominent method that overcomes this phenomen is the long short term memory which basicly consists of a memory that is controlled by two gate units which can be interpreted as adaptive encoding and recall units. Critical points which make the processing of the structure differ from human processing concern the encoding and the storage: (1) The structureis built to massively store information instead of carefully selecting few impressions for storage in memory.(2) Reweighting of stored information due to changing constellations is not possible. Coming back to the example this would mean that a nice action at the beginning of the game has a strong impact on your categorization independent of what kind of actions - might they be impressive or not - followed afterwards. In this work we tackle these points through introducing an attentionmechanism which drives the encoding and the storage of the structure. We analyse the model behavior in category learning tasks.

AB - Categorization is a skill which is used extensively in everyday life and as therefore an important aspect of human cognition. Consequently a variety of studies exist which address the topic and revealed that diverse factors affect human categorization performance. A critical but not extensively studied factor is time. Imagine watching a basketball game for 30 minutes. In this period of time plenty of actions will take place resulting in diverse impressions which make you afterwards categorize the game as interesting or boring. Such a categorization task is very similar to a time series classification task in the context of machine learning. In the field of machinelearning a phenomen called “vanishing gradient” is known which makes it generally hard to solve such a categorization task. A prominent method that overcomes this phenomen is the long short term memory which basicly consists of a memory that is controlled by two gate units which can be interpreted as adaptive encoding and recall units. Critical points which make the processing of the structure differ from human processing concern the encoding and the storage: (1) The structureis built to massively store information instead of carefully selecting few impressions for storage in memory.(2) Reweighting of stored information due to changing constellations is not possible. Coming back to the example this would mean that a nice action at the beginning of the game has a strong impact on your categorization independent of what kind of actions - might they be impressive or not - followed afterwards. In this work we tackle these points through introducing an attentionmechanism which drives the encoding and the storage of the structure. We analyse the model behavior in category learning tasks.

KW - LSTM

KW - Memory

KW - Attention

KW - Categorization

KW - Modelling

M3 - Conference contribution/Paper

SN - 0976831813

SP - 845

EP - 850

BT - CogSci 2005

A2 - Bara, Bruno G.

A2 - Barsalou, Lawrence

A2 - Bucciarelli, Monica

PB - Cognitive Science Society

ER -