Can Dimension of latent representation get bigger? For example, 10, 128

[yasuobinggan]

Thank you for your code!
Sorry, I don't know much about TF.
When I tried the unsupervised experiment, I tried to improve the dimension of latent representation(like, 10, 128), but the generator was completely broken, and I failed to try to modify it. I wonder if you have tried, could you please provide some information? Thank you very much!