The Role of Dopamine in the Prioritization of Information and Reward in Reinforcement Learning
Session Number
BHVSO 12
Advisor(s)
Akram Bakkour
Abby Shivers, University of Chicago
Discipline
Behavioral and Social Sciences
Start Date
17-4-2024 10:25 AM
End Date
17-4-2024 10:40 AM
Abstract
Extant literature suggests that rewards in the environment drive learning. Additionally, older adults have demonstrated an impaired ability to learn from reward feedback. However, recent work indicates that information can be just as, if not more, useful as rewards in reinforcement learning paradigms. The Battleship project aimed to explore how these cues uniquely guide learning about one’s environment differently across ages. Participants aged 18-57 completed a hidden shape search task inspired by the game Battleship. Participants had to infer the hidden shape on each trial. There were five shapes possible in the task. The time it took participants to learn the shape set was measured as well as the efficiency by which they searched for shapes.
Surprisingly, older individuals learned the shapes faster than younger participants. Furthermore, faster learning was associated with a more efficient search process based on gaining information. Thus, learning in this task is improved by a search process that is based on gaining information from one’s choices, which becomes more efficient as age increases. These results have the potential to inform our understanding of the cognitive mechanisms facilitating reinforcement learning and aid in developing interventions that optimize learning.
The Role of Dopamine in the Prioritization of Information and Reward in Reinforcement Learning
Extant literature suggests that rewards in the environment drive learning. Additionally, older adults have demonstrated an impaired ability to learn from reward feedback. However, recent work indicates that information can be just as, if not more, useful as rewards in reinforcement learning paradigms. The Battleship project aimed to explore how these cues uniquely guide learning about one’s environment differently across ages. Participants aged 18-57 completed a hidden shape search task inspired by the game Battleship. Participants had to infer the hidden shape on each trial. There were five shapes possible in the task. The time it took participants to learn the shape set was measured as well as the efficiency by which they searched for shapes.
Surprisingly, older individuals learned the shapes faster than younger participants. Furthermore, faster learning was associated with a more efficient search process based on gaining information. Thus, learning in this task is improved by a search process that is based on gaining information from one’s choices, which becomes more efficient as age increases. These results have the potential to inform our understanding of the cognitive mechanisms facilitating reinforcement learning and aid in developing interventions that optimize learning.