I’ve been splicing animals with Stable Diffusion (v1.2 in an attempt to get more wildlife images and less paintings) using spherical lerping of different animal name encodings. The two animals are listed in the captions / alt text.
I do enjoy how the fluffy chicken / toad looks just as grumpy as a toad.
Combination of Toad + Frog + Spider by mixing together the token embeddings for the 3 words within the language model for a diffusion text to image generative model.