Dynamic Interaction between Reinforcement Learning and Attention in Multidimensional Environments

详细信息    查看全文

• 作者：Yuan Chang Leong¹ ; ⁶ ; ^{ycleong@stanford.edu} ; Angela Radulescu² ; ⁶ ; ^{angelar@princeton.edu} ; Reka Daniel³ ; Vivian DeWoskin⁴ ; Yael Niv² ; ⁵ ; ⁷ ; ^{yael@princeton.edu}
• 关键词：reinforcement learning ; attention ; decision making ; computational modeling ; MVPA ; fMRI ; vmPFC ; striatum ; value ; prediction error
• 刊名：Neuron
• 出版年：2017
• 出版时间：18 January 2017
• 年：2017
• 卷：93
• 期：2
• 页码：451-463
• 全文大小：1943 K
• 卷排序：93

文摘

Attention constrains reinforcement learning processes to relevant task dimensions Learned values of stimulus features drive shifts in the focus of attention Attention biases value and reward prediction error signals in the brain Dynamic control of attention is associated with activity in frontoparietal network