Genetic variation created by whole genome duplication (WGD) events can provide beneficial genetic variation for evolution to act upon. However, polyploid genomes comprise many, sometimes unnecessary, duplicated genes. The loss of these duplicates is part of the process of diploidization. One mechanism of diploidization is fractionation, where genes are lost or silenced in one or both homoeologs. Fractionation structurally and functionally alters the genome as only a subset of duplicated genes is retained. This affects the amount of genetic variation present following diploidization, influencing how a lineage will be able to evolve. Despite its evolutionary importance, little is known about how fractionation occurs in different lineages of land plants, particularly in ferns and lycophytes (pteridophytes). In angiosperms, fractionation is thought to proceed quickly, mainly by gene loss. Previous work suggests that other land plants may not return to a diploid state as quickly or efficiently. In particular, ferns have more chromosomes and larger genomes than angiosperms, despite having the same or fewer WGD events. A few studies have found putatively silenced gene copies in ferns with a history of WGD. Additionally, a slower rate of chromosome loss has been found in ferns compared to angiosperms. We use newly available published genome sequences for ferns and lycophytes to examine how genome fractionation may be occurring. We leverage machine learning to identify the number of retained paleologs in species with a known history of WGD. Paleolog silencing or loss is then used to determine what type of fractionation is occurring following WGD in ferns and if there is evidence of biased gene retention across sub-genomes. Including pteridophytes, for the first time, in comparative work on fractionation is a key step to better understanding genome evolution in land plants.