(A) We simulate the trajectory of a bat exploring a volume of space over a prolonged period of time. At each step, the bat moves forward at a constant speed and chooses a new direction of movement close to the previous one. (B) The feed-forward network (equivalent to that shown in [Kropff and Treves, 2008]), including here only two mEC neurons. The would-be grid units receive inputs from place cell-like units with firing fields similar to those reported in Yartsev and Ulanovsky (2013). (C) Snapshots of the evolution of the firing rate map of a single unit. The figures correspond to (from left to right) 2, 5, 10 and 20 million time steps of learning.
© 2013, M Yartsev, N Ulanovsky. Figure 1 is reproduced from M Yartsev, N Ulanovsky. 2013. Representation of three-dimensional space in the hippocampus of flying bats. Science 340:367–372. Reprinted with permission from AAAS.