Working memory as a representational template for reinforcement learning
Working memory (WM) and reinforcement learning (RL) both influence decision-making, but how they interact to affect behaviour remains unclear. We assessed whether RL is influenced by the format of visual stimuli held in WM, either feature-based or unified, object-based representations. In a pre-regi...
| Main Authors: | , , , , |
|---|---|
| Format: | Journal article |
| Language: | English |
| Published: |
Nature Research
2024
|
| _version_ | 1817931397921243136 |
|---|---|
| author | Shibata, K Klar, V Fallon, SJ Husain, M Manohar, SG |
| author_facet | Shibata, K Klar, V Fallon, SJ Husain, M Manohar, SG |
| author_sort | Shibata, K |
| collection | OXFORD |
| description | Working memory (WM) and reinforcement learning (RL) both influence decision-making, but how they interact to affect behaviour remains unclear. We assessed whether RL is influenced by the format of visual stimuli held in WM, either feature-based or unified, object-based representations. In a pre-registered paradigm, participants learned stimulus-action combinations that provided reward through 80% probabilistic feedback. In parallel, participants retained the RL stimulus in WM and were asked to recall this stimulus after each RL choice. Crucially, the format of representation probed in WM was manipulated, with blocks encouraging either separate features or bound objects to be remembered. Incentivising a feature-based WM representation facilitated feature-based learning, shown by an improved choice strategy. This reveals a role of WM in providing sustained internal representations that are harnessed by RL, providing a framework by which these two cognitive processes cooperate. |
| first_indexed | 2024-12-09T03:21:23Z |
| format | Journal article |
| id | oxford-uuid:c25a35cb-1c6d-483a-acc4-d38780dd5285 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-12-09T03:21:23Z |
| publishDate | 2024 |
| publisher | Nature Research |
| record_format | dspace |
| spelling | oxford-uuid:c25a35cb-1c6d-483a-acc4-d38780dd52852024-11-12T20:10:33ZWorking memory as a representational template for reinforcement learningJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:c25a35cb-1c6d-483a-acc4-d38780dd5285EnglishJisc Publications RouterNature Research2024Shibata, KKlar, VFallon, SJHusain, MManohar, SGWorking memory (WM) and reinforcement learning (RL) both influence decision-making, but how they interact to affect behaviour remains unclear. We assessed whether RL is influenced by the format of visual stimuli held in WM, either feature-based or unified, object-based representations. In a pre-registered paradigm, participants learned stimulus-action combinations that provided reward through 80% probabilistic feedback. In parallel, participants retained the RL stimulus in WM and were asked to recall this stimulus after each RL choice. Crucially, the format of representation probed in WM was manipulated, with blocks encouraging either separate features or bound objects to be remembered. Incentivising a feature-based WM representation facilitated feature-based learning, shown by an improved choice strategy. This reveals a role of WM in providing sustained internal representations that are harnessed by RL, providing a framework by which these two cognitive processes cooperate. |
| spellingShingle | Shibata, K Klar, V Fallon, SJ Husain, M Manohar, SG Working memory as a representational template for reinforcement learning |
| title | Working memory as a representational template for reinforcement learning |
| title_full | Working memory as a representational template for reinforcement learning |
| title_fullStr | Working memory as a representational template for reinforcement learning |
| title_full_unstemmed | Working memory as a representational template for reinforcement learning |
| title_short | Working memory as a representational template for reinforcement learning |
| title_sort | working memory as a representational template for reinforcement learning |
| work_keys_str_mv | AT shibatak workingmemoryasarepresentationaltemplateforreinforcementlearning AT klarv workingmemoryasarepresentationaltemplateforreinforcementlearning AT fallonsj workingmemoryasarepresentationaltemplateforreinforcementlearning AT husainm workingmemoryasarepresentationaltemplateforreinforcementlearning AT manoharsg workingmemoryasarepresentationaltemplateforreinforcementlearning |