It had been shown experimentally in rats that, if a path is learned while awake, the place cells "replay" the path when the brain is showing sleep patterns. The cause was not entirely known, and this replay has been observed to follow the path through the place cells both in the forward and reverse directions. Molter et al hypothesized that the root of the replay was in LTP. They proposed that, during random brain activity during rest, a place cell could be activated, and then due to the LTP generated when the rat was awake, the rest of the cells along the path would be induced to spike sequentially.
Instead of experimenting with live animals, Molter decided to study a computer model of the system. He designed a program using custom self-made C++ tools. This program modeled the rat hippocampus as two layers, plus one layer of sensory input related to the space fields in the environment. This model will be described in greater detail later. Molter's program could simulate both 1D, in which the rat could only move forward along a straight-line path, and 2D, in which the rat could move along a free or constrained path in a two-dimensional room.