English

Fixing the problems of deep neural networks will require better training data and learning algorithms

Computer Vision and Pattern Recognition 2023-11-23 v1 Artificial Intelligence Machine Learning Neurons and Cognition

Abstract

Bowers and colleagues argue that DNNs are poor models of biological vision because they often learn to rival human accuracy by relying on strategies that differ markedly from those of humans. We show that this problem is worsening as DNNs are becoming larger-scale and increasingly more accurate, and prescribe methods for building DNNs that can reliably model biological vision.

Keywords

Cite

@article{arxiv.2311.12819,
  title  = {Fixing the problems of deep neural networks will require better training data and learning algorithms},
  author = {Drew Linsley and Thomas Serre},
  journal= {arXiv preprint arXiv:2311.12819},
  year   = {2023}
}

Comments

Published as a commentary in Behavioral and Brain Sciences

R2 v1 2026-06-28T13:27:43.161Z