CaltechAUTHORS
  A Caltech Library Service

Signed and unsigned reward prediction errors dynamically enhance learning and memory

Rouhani, Nina and Niv, Yael (2021) Signed and unsigned reward prediction errors dynamically enhance learning and memory. eLife, 10 . Art. No. e61077. ISSN 2050-084X. PMCID PMC8041467. doi:10.7554/eLife.61077. https://resolver.caltech.edu/CaltechAUTHORS:20210429-105343327

[img] PDF - Published Version
Creative Commons Attribution.

1MB
[img] MS Word (Transparent reporting form) - Supplemental Material
Creative Commons Attribution.

251kB

Use this Persistent URL to link to this item: https://resolver.caltech.edu/CaltechAUTHORS:20210429-105343327

Abstract

Memory helps guide behavior, but which experiences from the past are prioritized? Classic models of learning posit that events associated with unpredictable outcomes as well as, paradoxically, predictable outcomes, deploy more attention and learning for those events. Here, we test reinforcement learning and subsequent memory for those events, and treat signed and unsigned reward prediction errors (RPEs), experienced at the reward-predictive cue or reward outcome, as drivers of these two seemingly contradictory signals. By fitting reinforcement learning models to behavior, we find that both RPEs contribute to learning by modulating a dynamically changing learning rate. We further characterize the effects of these RPE signals on memory and show that both signed and unsigned RPEs enhance memory, in line with midbrain dopamine and locus-coeruleus modulation of hippocampal plasticity, thereby reconciling separate findings in the literature.


Item Type:Article
Related URLs:
URLURL TypeDescription
https://doi.org/10.7554/eLife.61077DOIArticle
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8041467PubMed CentralArticle
https://doi.org/10.7554/eLife.61077.sa1DOIDecision letter
https://doi.org/10.7554/eLife.61077.sa2DOIAuthor response
https://github.com/ninarouhani/2021_RouhaniNivRelated ItemCode
https://archive.softwareheritage.org/swh:1:rev:fa15d035dc4033ebad03f48dbd5c75b0c4d76c40Related ItemCode
ORCID:
AuthorORCID
Rouhani, Nina0000-0003-2814-0462
Niv, Yael0000-0002-0259-8371
Additional Information:© 2021 Rouhani and Niv. This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited. Received: 14 July 2020; Accepted: 26 February 2021; Published: 04 March 2021. We thank Angela Radulescu and Isabel Berwian for helpful comments. This work was supported by grant W911NF-14-1-0101 from the Army Research Office (YN), grant R01MH098861 from the National Institute for Mental Health (YN), grant R21MH120798 from the National Institute of Health (YN) and the National Science Foundation’s Graduate Research Fellowship Program (NR). The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication. Data availability: All data files and code for models, analysis and figures are publicly available at https://github.com/ninarouhani/2021_RouhaniNiv copy archived at https://archive.softwareheritage.org/swh:1:rev:fa15d035dc4033ebad03f48dbd5c75b0c4d76c40/. Author contributions: Nina Rouhani, Conceptualization, Data curation, Software, Formal analysis, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing - original draft, Project administration, Writing - review and editing; Yael Niv, Conceptualization, Resources, Supervision, Funding acquisition, Validation, Methodology, Writing - original draft, Writing - review and editing. Ethics: Human subjects: We obtained informed consent online; procedures were approved by Princeton University’s Institutional Review Board (IRB #4452).
Funders:
Funding AgencyGrant Number
Army Research Office (ARO)W911NF-14-1-0101
NIHR01MH098861
NIHR21MH120798
NSF Graduate Research FellowshipUNSPECIFIED
PubMed Central ID:PMC8041467
DOI:10.7554/eLife.61077
Record Number:CaltechAUTHORS:20210429-105343327
Persistent URL:https://resolver.caltech.edu/CaltechAUTHORS:20210429-105343327
Official Citation:Signed and unsigned reward prediction errors dynamically enhance learning and memory. Rouhani and Niv. eLife 2021;10:e61077. DOI: https://doi.org/10.7554/eLife.61077
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:108867
Collection:CaltechAUTHORS
Deposited By: Tony Diaz
Deposited On:29 Apr 2021 18:23
Last Modified:29 Apr 2021 18:23

Repository Staff Only: item control page