GLU is a text-based hybrid intensional language for programming conventional parallel computers whose cryptic, linear syntax often obscures static data and dimensional dependencies, thus making program functionality less apparent. We describe two different graphical models for GLU programming - an intensional model based on dependency graphs and an extensional model based on spread sheets - designed to better present data and dimensional dependencies in GLU programs. We compare the two models in their respective approaches for reveal ing latent data and dimensional dependencies in GLU programs.

• islip95-revised.ps