Replay in minds and machines

NEUROSCIENCE AND BIOBEHAVIORAL REVIEWS

卷:129

Article

Wittkuhn, Lennart^1,2  Chien, Samson^1,2  Hall-McMaster, Sam^1,2  Schuck, Nicolas W.^1,2 

[1] Max Planck Inst Human Dev, Max Planck Res Grp NeuroCode, Lentzeallee 94, D-14195 Berlin, Germany

[2] Max Planck UCL Ctr Computat Psychiat & Ageing Res, Lentzeallee 94, D-14195 Berlin, Germany

关键词: Replay;    Reinforcement learning;    Machine learning;    Representation learning;    Decision-making;   

DOI  :  10.1016/j.neubiorev.2021.08.002

来源: Elsevier

【 摘 要 】

Experience-related brain activity patterns reactivate during sleep, wakeful rest, and brief pauses from active behavior. In parallel, machine learning research has found that experience replay can lead to substantial performance improvements in artificial agents. Together, these lines of research suggest that replay has a variety of computational benefits for decision-making and learning. Here, we provide an overview of putative computational functions of replay as suggested by machine learning and neuroscientific research. We show that replay can lead to faster learning, less forgetting, reorganization or augmentation of experiences, and support planning and generalization. In addition, we highlight the benefits of reactivating abstracted internal representations rather than veridical memories, and discuss how replay could provide a mechanism to build internal representations that improve learning and decision-making.

【 授权许可】

Free