Replay in deep learning: current approaches and missing biological elements

DOI10.1162/NECO_A_01433MaRDI QIDQ5034463zbMATH OpenOpenAlexFDO

Authors Tyler L. Hayes, Giri Krishnan, Maxim Bazhenov, Hava T. Siegelmann, Terrence J. Sejnowski, Christopher Kanan

Publication date 18 February 2022

Published in Neural Computation (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2104.04132

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Neural networks for/in biological studies, artificial life and related topics (92B20)

Abstract: Replay is the reactivation of one or more neural patterns, which are similar to the activation patterns experienced during past waking experiences. Replay was first observed in biological neural networks during sleep, and it is now thought to play a critical role in memory formation, retrieval, and consolidation. Replay-like mechanisms have been incorporated into deep artificial neural networks that learn over time to avoid catastrophic forgetting of previous knowledge. Replay algorithms have been successfully used in a wide range of deep learning methods within supervised, unsupervised, and reinforcement learning paradigms. In this paper, we provide the first comprehensive comparison between replay in the mammalian brain and replay in artificial neural networks. We identify multiple aspects of biological replay that are missing in deep learning systems and hypothesize how they could be utilized to improve artificial neural networks.

Recommendations

• Hippocampal replay contributes to within session learning in a temporal difference reinforcement learning model
• AN ANALYSIS OF EXPERIENCE REPLAY IN TEMPORAL DIFFERENCE LEARNING
• How neural are neural networks? A comparison of information processing and storage in artificial and real neural systems
• Overcoming catastrophic forgetting in neural networks
• Biological grounding of recruitment Learning and vicinal algorithms in long-term potentiation

Cited in

(2)

• Methods for assessment of memory reactivation
• Saving memory space in deep neural networks by recomputing: a survey

This page was built for publication: Replay in deep learning: current approaches and missing biological elements

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5034463)