Training recurrent networks to generate hypotheses about how the brain solves hard navigation problems

Open in new window