OSCR

Competing value signals impair reward-learning via dopaminergic mechanisms and increase exploration.

Code ↔ Paper

The paper beside its authors' code: matches between them have not been computed for this paper yet.

Paper

Loaded from Europe PMC by your browser; OSCR does not store it: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

Stan · 138 lines · 3.9 KB · CC-BY-4.0

  1. functions {
  2. real partial_log_lik(array[,] int choice_slice,
  3. int start, int end,
  4. array[,] int reward,
  5. array[] real persev,
  6. array[] real alpha,
  7. array[] real beta,
  8. int nTrials) {
  9. real lp = 0;
  10. for (s in start:end) {
  11. vector[2] v = rep_vector(0.0, 2);
  12. vector[2] pb;
  13. for (t in 1:nTrials) {
  14. pb = rep_vector(0.0, 2);
  15. if (t > 1) {
  16. pb[ choice_slice[s, t-1] ] = persev[s];
  17. }
  18. lp += categorical_logit_lpmf(
  19. choice_slice[s, t] |
  20. beta[s] * (v + pb)
  21. );
  22. real pe = reward[s, t] - v[ choice_slice[s, t] ];
  23. v[ choice_slice[s, t] ] += alpha[s] * pe;
  24. }
  25. }
  26. return lp;
  27. }
  28. }
  29. data {
  30. int<lower=1> nSubjects;
  31. int<lower=1> nTrials;
  32. int<lower=1> nConditions;
  33. int<lower=1> nSessions;
  34. array[nSessions, nConditions, nSubjects, nTrials] int<lower=1, upper=2> choice;
  35. array[nSessions, nConditions, nSubjects, nTrials] int<lower=-1, upper=1> reward;
  36. }
  37. parameters {
  38. array[nSessions, nConditions] real persev_mu;
  39. array[nSessions, nConditions] real<lower=0> persev_sd;
  40. array[nSessions, nConditions] real<lower=0, upper=1> alpha_mu;
  41. array[nSessions, nConditions] real<lower=0> alpha_kappa;
  42. array[nSessions, nConditions] real beta_mu_log;
  43. array[nSessions, nConditions] real<lower=0> beta_sd_log;
  44. array[nSessions, nConditions, nSubjects] real persev_z;
  45. array[nSessions, nConditions, nSubjects] real beta_z;
  46. array[nSessions, nConditions, nSubjects] real<lower=0, upper=1> alpha;
  47. }
  48. transformed parameters {
  49. array[nSessions, nConditions, nSubjects] real persev;
  50. array[nSessions, nConditions, nSubjects] real<lower=0> beta;
  51. array[nSessions, nConditions] real<lower=0> alpha_a;
  52. array[nSessions, nConditions] real<lower=0> alpha_b;
  53. for (m in 1:nSessions) {
  54. for (c in 1:nConditions) {
  55. alpha_a[m, c] = alpha_mu[m, c] * alpha_kappa[m, c];
  56. alpha_b[m, c] = (1 - alpha_mu[m, c]) * alpha_kappa[m, c];
  57. for (s in 1:nSubjects) {
  58. persev[m, c, s] = persev_mu[m, c] + persev_sd[m, c] * persev_z[m, c, s];
  59. real log_beta = beta_mu_log[m, c] + beta_sd_log[m, c] * beta_z[m, c, s];
  60. beta[m, c, s] = exp(log_beta);
  61. }
  62. }
  63. }
  64. }
  65. model {
  66. for (m in 1:nSessions) {
  67. for (c in 1:nConditions) {
  68. alpha_mu[m, c] ~ beta(2, 2);
  69. alpha_kappa[m, c] ~ lognormal(log(10), 0.25);
  70. beta_mu_log[m, c] ~ normal(log(3), 0.25);
  71. beta_sd_log[m, c] ~ normal(0.2, 0.1) T[0, ];
  72. persev_mu[m, c] ~ normal(0.0, 0.5);
  73. persev_sd[m, c] ~ normal(0.5, 0.25) T[0, ];
  74. for (s in 1:nSubjects) {
  75. persev_z[m, c, s] ~ normal(0, 1);
  76. beta_z[m, c, s] ~ normal(0, 1);
  77. alpha[m, c, s] ~ beta(alpha_a[m, c], alpha_b[m, c]);
  78. }
  79. target += reduce_sum(
  80. partial_log_lik,
  81. choice[m, c],
  82. 8,
  83. reward[m, c],
  84. persev[m, c],
  85. alpha[m, c],
  86. beta[m, c],
  87. nTrials
  88. );
  89. }
  90. }
  91. }
  92. generated quantities {
  93. array[nSessions, nConditions, nSubjects, nTrials] real log_lik;
  94. array[nSessions, nConditions, nSubjects, nTrials, 2] real ev;
  95. array[nSessions, nConditions, nSubjects, nTrials] real pe;
  96. for (m in 1:nSessions) {
  97. for (c in 1:nConditions) {
  98. for (s in 1:nSubjects) {
  99. vector[2] v = rep_vector(0.0, 2);
  100. vector[2] pb;
  101. for (t in 1:nTrials) {
  102. pb = rep_vector(0.0, 2);
  103. if (t > 1) {
  104. pb[choice[m, c, s, t-1]] = persev[m, c, s];
  105. }
  106. log_lik[m, c, s, t] = categorical_logit_lpmf(
  107. choice[m, c, s, t] |
  108. beta[m, c, s] * (v + pb)
  109. );
  110. ev[m, c, s, t, 1] = v[1];
  111. ev[m, c, s, t, 2] = v[2];
  112. pe[m, c, s, t] = reward[m, c, s, t] - v[ choice[m, c, s, t] ];
  113. v[ choice[m, c, s, t] ] += alpha[m, c, s] * pe[m, c, s, t];
  114. }
  115. }
  116. }
  117. }
  118. }

pers.stan, under CC-BY-4.0 · at the source

Overview

Authors: Wen-Wei Lin1, Pei-Yu Lee1, Hsin-Yun Tsai2, Yi-Hsuan Lin2, Min-Min Lin1, Zheng-Liang Lu3, Mei-Yu Yeh1, Ming-Tsung Tseng1
ORCID iDs: Ming-Tsung Tseng
  1. Graduate Institute of Brain and Mind Sciences, National Taiwan University College of Medicine, Taipei, Taiwan
  2. Taiwan International Graduate Program in Interdisciplinary Neuroscience, National Taiwan University and Academia Sinica, Taipei, Taiwan
  3. Department of Computer Science and Information Engineering, National Taiwan University, Taipei, Taiwan
Institutions: National Taiwan University (Taiwan); Academia Sinica (Taiwan)
Journal: PLoS biology, volume 24, issue 7, article e3003922
Dates: received 9 February 2026; accepted 13 July 2026; published online 24 July 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.1371/journal.pbio.3003922 · PMID 42497225 · PMCID PMC13427004 · OpenAlex W7170303932
Open access: gold, a free copy (OpenAlex)
Status: code verified (how the code is found and verified)
Categories: human (organism), cognitive (subfield)
Methods: Statistics, Machine learning, Connectivity, fMRI & imaging
MeSH: Dopamine*, Learning*, Reward*, Amisulpride, Animals, Choice Behavior, Conditioning, Operant, Female, Humans, Magnetic Resonance Imaging, Male, Punishment, Reinforcement Machine Learning, Reinforcement, Psychology, Sulpiride (* major topic)
Topic: Neural and Behavioral Psychology Studies (Cognitive Neuroscience, Neuroscience), according to OpenAlex
Funding: National Science and Technology Council (NSTC 113-2314-B-002-273-MY3); Ministry of Science and Technology of Taiwan (MOST 107-2314-B-002-073 and MOST 108-2410-H-002-110-MY2)
Citations: not cited yet (Europe PMC); 105 references in the paper

Abstract

Effective reinforcement learning requires balancing exploration of uncertain options with exploitation of known outcomes. In real-world contexts, the same action may yield rewards in some situations and punishments in others, yet how these learning processes influence each other remains unclear. Here, we examine the neural mechanisms underlying how reward and punishment learning interact to guide adaptive behavior. We conducted four experiments (N = 159) using an instrumental learning task with binary choices, some of which were exclusive to reward or punishment learning trials, while others appeared in both, allowing assessment of their interaction. When choices were tied to a single learning type, reward learning engages less exploration (i.e., fewer choices of lower-value options) than punishment learning. Critically, when both learning processes were concurrently engaged, reward learning was selectively impaired, accompanied by enhanced exploration and greater activation in exploration-related prefrontal regions revealed by fMRI. Computational modeling showed that impaired reward learning was best explained by sensitivity to prior punishment history associated with reward-learning options, while individual differences in loss aversion predicted the degree of increased exploration. Finally, pharmacological attenuation of dopaminergic signaling via the D2/3 receptor antagonist amisulpride abolished both the increased exploration and the interference with reward learning. These findings suggest that punishment-history interference during reward learning is dopamine-modulated and associated with increased exploration, with individual differences in this exploration linked to loss aversion, providing a mechanistic account of how the brain resolves competing value signals and informing dopamine-related learning disturbances in neuropsychiatric conditions.

Reproduced under the paper's license (CC BY), from the paper cited above.

Repositories

Its files are read in the Code ↔ Paper reader above.

Zenodo 21261694

License: CC-BY-4.0
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Size: 1 file
Software Heritage: not checked
Found in: “Data Availability”
Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: Stan (6 files)
Availability: 1 check, the latest on 27 September 2026: the link answers (HTTP 200)
  • 27 September 2026: the link answers (HTTP 200)
8 files

linwenwei6915/reward-punishment-interference-stan-models

License: MIT
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Commit: 0f3f66dff9fbfc00e9ce123d723e970293af7dab, 8 July 2026
Languages: Stan (6)
Size: 8 files, 6 scripts
Software Heritage: not archived
Found in: the Zenodo archive record
Holds: README, license file
Not found: CITATION.cff, environment file, tests, continuous integration, documentation
Tools: Stan (6 files)
Availability: 1 check, the latest on 27 September 2026: the link answers
  • 27 September 2026: the link answers
8 files

The paper's code and data availability statement is in the Data section.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 2 repositories of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 12 scripts, each with its path and the digest of its content;
  • no match between paragraphs and code yet;
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

Datasets cited

Data Availability

De-identified behavioral data, the numerical data underlying the figure plots (S1 Data), and unthresholded fMRI statistical maps are available on the Open Science Framework (https://doi.org/10.17605/OSF.IO/T7YWA). The dataset and analysis scripts are available on Zenodo (https://doi.org/10.5281/zenodo.21261694), with a README describing the repository contents.

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 8 authors, 15 MeSH terms, 2 funders, 92 references.

Cite

This paper

Lin, W.-W., Lee, P.-Y., Tsai, H.-Y., Lin, Y.-H., Lin, M.-M., Lu, Z.-L., Yeh, M.-Y., & Tseng, M.-T. (2026). Competing value signals impair reward-learning via dopaminergic mechanisms and increase exploration. PLoS biology, 24(7), e3003922. https://doi.org/10.1371/journal.pbio.3003922

BibTeX

@article{lin2026competing,
author = {Lin, Wen-Wei and Lee, Pei-Yu and Tsai, Hsin-Yun and Lin, Yi-Hsuan and Lin, Min-Min and Lu, Zheng-Liang and Yeh, Mei-Yu and Tseng, Ming-Tsung},
title = {{Competing value signals impair reward-learning via dopaminergic mechanisms and increase exploration}},
journal = {PLoS biology},
year = {2026},
month = jul,
volume = {24},
number = {7},
pages = {e3003922},
publisher = {PLOS},
issn = {1544-9173},
doi = {10.1371/journal.pbio.3003922},
url = {https://doi.org/10.1371/journal.pbio.3003922},
pmid = {42497225},
pmcid = {PMC13427004}
}

RIS

TY - JOUR
AU - Lin, Wen-Wei
AU - Lee, Pei-Yu
AU - Tsai, Hsin-Yun
AU - Lin, Yi-Hsuan
AU - Lin, Min-Min
AU - Lu, Zheng-Liang
AU - Yeh, Mei-Yu
AU - Tseng, Ming-Tsung
TI - Competing value signals impair reward-learning via dopaminergic mechanisms and increase exploration
T2 - PLoS biology
J2 - PLoS Biol
PY - 2026
DA - 2026/07/24
VL - 24
IS - 7
SP - e3003922
SN - 1544-9173
PB - PLOS
DO - 10.1371/journal.pbio.3003922
UR - https://doi.org/10.1371/journal.pbio.3003922
LA - en
ER -

CSL-JSON

{
"id": "10.1371/journal.pbio.3003922",
"type": "article-journal",
"title": "Competing value signals impair reward-learning via dopaminergic mechanisms and increase exploration",
"container-title": "PLoS biology",
"author": [
{
"family": "Lin",
"given": "Wen-Wei"
},
{
"family": "Lee",
"given": "Pei-Yu"
},
{
"family": "Tsai",
"given": "Hsin-Yun"
},
{
"family": "Lin",
"given": "Yi-Hsuan"
},
{
"family": "Lin",
"given": "Min-Min"
},
{
"family": "Lu",
"given": "Zheng-Liang"
},
{
"family": "Yeh",
"given": "Mei-Yu"
},
{
"family": "Tseng",
"given": "Ming-Tsung"
}
],
"container-title-short": "PLoS Biol",
"volume": "24",
"issue": "7",
"page": "e3003922",
"DOI": "10.1371/journal.pbio.3003922",
"PMID": "42497225",
"PMCID": "PMC13427004",
"ISSN": "1544-9173",
"publisher": "PLOS",
"URL": "https://doi.org/10.1371/journal.pbio.3003922",
"language": "en",
"issued": {
"date-parts": [
[
2026,
7,
24
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.1007/s42113-024-00202-y [code]
Exploration-Exploitation Mechanisms in Recurrent Neural Networks and Human Learners in Restless Bandit Problems.
Journal: Computational brain & behavior
In common: Stan, 15 references
[2] doi:10.1002/hbm.25988
Reinforcement learning with associative or discriminative generalization across states and actions: fMRI at 3 T and 7 T.
Journal: Human brain mapping
In common: cognitive, 16 references
[3] doi:10.1038/s41598-024-76025-5 [code]
A causal role of the right dorsolateral prefrontal cortex in random exploration.
Journal: Scientific reports
In common: 12 references
[4] doi:10.1371/journal.pbio.3002031 [code]
Individuals with problem gambling and obsessive-compulsive disorder learn through distinct reinforcement mechanisms.
Journal: PLoS biology
In common: cognitive, 10 references
[5] doi:10.7554/elife.74149
Different brain systems support learning from received and avoided pain during human pain-avoidance learning.
Journal: eLife
In common: cognitive, 9 references
[6] doi:10.1093/brain/awae025 [code]
Impaired value-based decision-making in Parkinson's disease apathy.
Journal: Brain : a journal of neurology
In common: Stan, cognitive, 8 references
[7] doi:10.1038/s41598-024-64240-z [code]
Uncertainty of treatment efficacy moderates placebo effects on reinforcement learning.
Journal: Scientific reports
In common: Stan, 8 references
[8] doi:10.1126/sciadv.adi4927 [code]
Subsecond fluctuations in extracellular dopamine encode reward and punishment prediction errors in humans.
Journal: Science advances
In common: Stan, cognitive, 7 references
[9] doi:10.1038/s41562-025-02324-0 [code]
Hybrid neural-cognitive models reveal how memory shapes human reward learning.
Journal: Nature human behaviour
In common: cognitive, 8 references
[10] doi:10.1007/s42113-025-00258-4 [code]
Recurrent Neural Network Exploration Strategies During Reinforcement Learning Depend on Network Capacity.
Journal: Computational brain & behavior
In common: Stan, 7 references

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Add the badge to the code's README

OSCR: the badge of this paper

If you are an author or a maintainer of this paper's code, you can add this badge. It says what OSCR knows of the paper (its code, its validated tracing map) and links the code to this page. Paste one of these snippets under the README's title. Keep the two marks around it, so that the badge can be replaced in place and is never added twice. OSCR changes nothing in your repository.

Markdown
<!-- oscr-badge:start -->
[![OSCR](https://openscicode.org/badge/doi/10.1371/journal.pbio.3003922.svg)](https://openscicode.org/badge/doi/10.1371/journal.pbio.3003922)
<!-- oscr-badge:end -->
reStructuredText
.. oscr-badge:start

.. image:: https://openscicode.org/badge/doi/10.1371/journal.pbio.3003922.svg
   :target: https://openscicode.org/badge/doi/10.1371/journal.pbio.3003922
   :alt: OSCR

.. oscr-badge:end
HTML
<!-- oscr-badge:start -->
<a href="https://openscicode.org/badge/doi/10.1371/journal.pbio.3003922"><img src="https://openscicode.org/badge/doi/10.1371/journal.pbio.3003922.svg" alt="OSCR"></a>
<!-- oscr-badge:end -->

On GitHub, open the README, choose Edit (the pencil), paste the snippet under the title, then "Commit changes".

If you choose to, you can have the badge kept up to date for you, with oscr badge --write from the command line or with the GitHub Action in your repository's workflows. Every other form and option (a card, AsciiDoc, LaTeX, shields.io) is on the badge page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: code errors and mismatches between the code and the paper about this paper and its code, from readers and its authors, beside the paper and its tracing map.

Report a mismatch between the code and the paper · Report a code error (signed in)

Reproductions: reports from readers who ran the authors' code and did not get the paper's result, what they ran, with which environment, commit and data.

Report a reproduction failure (signed in)

Versions of its code: the releases of its code tied to a version of this paper (the preprint, the accepted manuscript, the version of record…), each with the tracing map versioned with it.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.