BinaryRBM gets very noisy fantasy particles

I ran through the MNIST BinaryRBM example with no modifications. The output fantasy particles are noisy and lack variability. Then I checked the output in the MNIST example and see that the same effects are there too. Noisy and no variability. <img width="642" alt="pcd_bin" src="https://github.com/cossio/RestrictedBoltzmannMachines.jl/assets/1161170/7617846b-e712-464e-a155-9727f613aeed"> Next, I tried the fast_pcd, which is supposedly an improved sampling method. But it's even worse <img width="634" alt="fast_pcd" src="https://github.com/cossio/RestrictedBoltzmannMachines.jl/assets/1161170/e97e1873-fc97-4420-991c-6e14770cb9c2"> Is there something that can be done to achieve better results? This is really not very useful as is.