For Human-Like Models, Train on Human-Like Tasks Journal Article uri icon



  • Bowers et al. (2022) express skepticism about deep neural networks (DNNs) as models of human vision due to DNNs’ failures to account for results from psychological research. We argue that to fairly assess DNNs, we must first train them on more human-like tasks which we hypothesize will induce more human-like behaviors and representations.

publication date

  • March 2, 2023

has restriction

  • green

Date in CU Experts

  • March 14, 2023 1:09 AM

Full Author List

  • Hermann K; Nayebi A; van Steenkiste S; Jones M

author count

  • 4

Other Profiles