The relative response of neutron rem detectors has previously been shown to increase after prolonged exposure to a neutron flux. This increase has been referred to as the “soak factor.” The cause of the increased response has been previously unexplained. This note reports on a search for the underlying cause of the increased response. Testing involved gamma-ray spectroscopy of activated neutron rem detector components and testing of instrument response at various rates of neutron flux and accumulated fluence. The proposed primary cause of the increased response is activation of copper in the brass collar at the fill end of the gas-filled detector tube. This component allows for attachment of the extension housing on the tube containing BF3 or 3He gas. Under a neutron flux, some of this copper activates to 64Cu. In addition, 56Mn was detected in activated components but does not seem to be a significant contributor to the detector response. Cadmium isotopes did not appear to be significant. The activated component causes an increase in indicated neutron dose rate due to decay photons from activated components. Photons normally have little impact on neutron rem detector readings because the electrical pulses produced in the probes are below the lower-level discriminator. However, the photon pulses do impact the overall count rate when they occur simultaneously with normally uncounted, lower amplitude, wall-effect neutron pulses.