Qasim Salman E, Deswal Aarushi, Saez Ignacio, Gu Xiaosi
Department of Psychiatry, Icahn School of Medicine at Mount Sinai, New York, NY, USA.
Center for Computational Psychiatry, Icahn School of Medicine at Mount Sinai, New York, NY, USA.
Commun Psychol. 2024 Jun 5;2(1):52. doi: 10.1038/s44271-024-00106-4.
How our decisions impact our memories is not well understood. Reward prediction errors (RPEs), the difference between expected and obtained reward, help us learn to make optimal decisions-providing a signal that may influence subsequent memory. To measure this influence and how it might go awry in mood disorders, we recruited a large cohort of human participants to perform a decision-making task in which perceptually memorable stimuli were associated with probabilistic rewards, followed by a recognition test for those stimuli. Computational modeling revealed that positive RPEs enhanced both the accuracy of memory and the temporal efficiency of memory search, beyond the contribution of perceptual information. Critically, positive affect upregulated the beneficial effect of RPEs on memory. These findings demonstrate how affect selectively regulates the impact of RPEs on memory, providing a computational mechanism for biased memory in mood disorders.
我们的决策如何影响记忆，目前还不太清楚。奖励预测误差（RPEs），即预期奖励与实际获得奖励之间的差异，帮助我们学会做出最优决策——提供一个可能影响后续记忆的信号。为了衡量这种影响以及它在情绪障碍中可能出现的偏差，我们招募了一大群人类参与者来执行一项决策任务，在该任务中，在感知上令人难忘的刺激与概率奖励相关联，随后对这些刺激进行识别测试。计算模型表明，积极的奖励预测误差提高了记忆的准确性和记忆搜索的时间效率，超出了感知信息的作用。至关重要的是，积极情绪上调了奖励预测误差对记忆的有益影响。这些发现表明情绪如何选择性地调节奖励预测误差对记忆的影响，为情绪障碍中的记忆偏差提供了一种计算机制。
