It does make sense though. Accurate global illumination is very strongly represented in nearly all training data (except illustrations) so it makes sense that t

It does make sense though. Accurate global illumination is very strongly represented in nearly all training data (except illustrations) so it makes sense that the model learned an approximation of it.