How glitter relates to gold

Similarity-dependent reward prediction errors in the human striatum

T. Kahnt, S.Q. Park, C. Burke, P.N. Tobler

Research output: Contribution to journalArticleScientificpeer-review

Abstract

Optimal choices benefit from previous learning. However, it is not clear how previously learned stimuli influence behavior to novel but similar stimuli. One possibility is to generalize based on the similarity between learned and current stimuli. Here, we use neuroscientific methods and a novel computational model to inform the question of how stimulus generalization is implemented in the human brain. Behavioral responses during an intradimensional discrimination task showed similarity-dependent generalization. Moreover, a peak shift occurred, i.e., the peak of the behavioral generalization gradient was displaced from the rewarded conditioned stimulus in the direction away from the unrewarded conditioned stimulus. To account for the behavioral responses, we designed a similarity-based reinforcement learning model wherein prediction errors generalize across similar stimuli and update their value. We show that this model predicts a similarity-dependent neural generalization gradient in the striatum as well as changes in responding during extinction. Moreover, across subjects, the width of generalization was negatively correlated with functional connectivity between the striatum and the hippocampus. This result suggests that hippocampus–striatal connections contribute to stimulus-specific value updating by controlling the width of generalization. In summary, our results shed light onto the neurobiology of a fundamental, similarity-dependent learning principle that allows learning the value of stimuli that have never been encountered.
Original languageEnglish
Pages (from-to)16521-16529
JournalJournal of Neuroscience
Volume32
Publication statusPublished - 2012

Fingerprint

Gold
Stimulus Generalization
Neurobiology

Cite this

Kahnt, T., Park, S. Q., Burke, C., & Tobler, P. N. (2012). How glitter relates to gold: Similarity-dependent reward prediction errors in the human striatum. Journal of Neuroscience, 32, 16521-16529.
Kahnt, T. ; Park, S.Q. ; Burke, C. ; Tobler, P.N. / How glitter relates to gold : Similarity-dependent reward prediction errors in the human striatum. In: Journal of Neuroscience. 2012 ; Vol. 32. pp. 16521-16529.
@article{5d7f373ad52d45a5a958549edc14424a,
title = "How glitter relates to gold: Similarity-dependent reward prediction errors in the human striatum",
abstract = "Optimal choices benefit from previous learning. However, it is not clear how previously learned stimuli influence behavior to novel but similar stimuli. One possibility is to generalize based on the similarity between learned and current stimuli. Here, we use neuroscientific methods and a novel computational model to inform the question of how stimulus generalization is implemented in the human brain. Behavioral responses during an intradimensional discrimination task showed similarity-dependent generalization. Moreover, a peak shift occurred, i.e., the peak of the behavioral generalization gradient was displaced from the rewarded conditioned stimulus in the direction away from the unrewarded conditioned stimulus. To account for the behavioral responses, we designed a similarity-based reinforcement learning model wherein prediction errors generalize across similar stimuli and update their value. We show that this model predicts a similarity-dependent neural generalization gradient in the striatum as well as changes in responding during extinction. Moreover, across subjects, the width of generalization was negatively correlated with functional connectivity between the striatum and the hippocampus. This result suggests that hippocampus–striatal connections contribute to stimulus-specific value updating by controlling the width of generalization. In summary, our results shed light onto the neurobiology of a fundamental, similarity-dependent learning principle that allows learning the value of stimuli that have never been encountered.",
author = "T. Kahnt and S.Q. Park and C. Burke and P.N. Tobler",
year = "2012",
language = "English",
volume = "32",
pages = "16521--16529",
journal = "Journal of Neuroscience",
issn = "0270-6474",
publisher = "SOC NEUROSCIENCE",

}

How glitter relates to gold : Similarity-dependent reward prediction errors in the human striatum. / Kahnt, T.; Park, S.Q.; Burke, C.; Tobler, P.N.

In: Journal of Neuroscience, Vol. 32, 2012, p. 16521-16529.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - How glitter relates to gold

T2 - Similarity-dependent reward prediction errors in the human striatum

AU - Kahnt, T.

AU - Park, S.Q.

AU - Burke, C.

AU - Tobler, P.N.

PY - 2012

Y1 - 2012

N2 - Optimal choices benefit from previous learning. However, it is not clear how previously learned stimuli influence behavior to novel but similar stimuli. One possibility is to generalize based on the similarity between learned and current stimuli. Here, we use neuroscientific methods and a novel computational model to inform the question of how stimulus generalization is implemented in the human brain. Behavioral responses during an intradimensional discrimination task showed similarity-dependent generalization. Moreover, a peak shift occurred, i.e., the peak of the behavioral generalization gradient was displaced from the rewarded conditioned stimulus in the direction away from the unrewarded conditioned stimulus. To account for the behavioral responses, we designed a similarity-based reinforcement learning model wherein prediction errors generalize across similar stimuli and update their value. We show that this model predicts a similarity-dependent neural generalization gradient in the striatum as well as changes in responding during extinction. Moreover, across subjects, the width of generalization was negatively correlated with functional connectivity between the striatum and the hippocampus. This result suggests that hippocampus–striatal connections contribute to stimulus-specific value updating by controlling the width of generalization. In summary, our results shed light onto the neurobiology of a fundamental, similarity-dependent learning principle that allows learning the value of stimuli that have never been encountered.

AB - Optimal choices benefit from previous learning. However, it is not clear how previously learned stimuli influence behavior to novel but similar stimuli. One possibility is to generalize based on the similarity between learned and current stimuli. Here, we use neuroscientific methods and a novel computational model to inform the question of how stimulus generalization is implemented in the human brain. Behavioral responses during an intradimensional discrimination task showed similarity-dependent generalization. Moreover, a peak shift occurred, i.e., the peak of the behavioral generalization gradient was displaced from the rewarded conditioned stimulus in the direction away from the unrewarded conditioned stimulus. To account for the behavioral responses, we designed a similarity-based reinforcement learning model wherein prediction errors generalize across similar stimuli and update their value. We show that this model predicts a similarity-dependent neural generalization gradient in the striatum as well as changes in responding during extinction. Moreover, across subjects, the width of generalization was negatively correlated with functional connectivity between the striatum and the hippocampus. This result suggests that hippocampus–striatal connections contribute to stimulus-specific value updating by controlling the width of generalization. In summary, our results shed light onto the neurobiology of a fundamental, similarity-dependent learning principle that allows learning the value of stimuli that have never been encountered.

M3 - Article

VL - 32

SP - 16521

EP - 16529

JO - Journal of Neuroscience

JF - Journal of Neuroscience

SN - 0270-6474

ER -