Current State and Future Directions for Learning in Biological Recurrent Neural Networks: A Perspective Piece
Neurons and Cognition
2022-01-06 v2 Artificial Intelligence
Abstract
We provide a brief review of the common assumptions about biological learning with findings from experimental neuroscience and contrast them with the efficiency of gradient-based learning in recurrent neural networks. The key issues discussed in this review include: synaptic plasticity, neural circuits, theory-experiment divide, and objective functions. We conclude with recommendations for both theoretical and experimental neuroscientists when designing new studies that could help bring clarity to these issues.
Keywords
Cite
@article{arxiv.2105.05382,
title = {Current State and Future Directions for Learning in Biological Recurrent Neural Networks: A Perspective Piece},
author = {Luke Y. Prince and Roy Henha Eyono and Ellen Boven and Arna Ghosh and Joe Pemberton and Franz Scherr and Claudia Clopath and Rui Ponte Costa and Wolfgang Maass and Blake A. Richards and Cristina Savin and Katharina Anna Wilmes},
journal= {arXiv preprint arXiv:2105.05382},
year = {2022}
}