These works document post-natural specimens produced through a two-stage experimental process: procedural algorithms generate initial forms, which are then processed through fine-tuned diffusion models. The resulting organisms exhibit morphological logics that neither human intention nor machine training could produce independently, and because they exist only in computational substrates, the images themselves function as both experimental results and archival documentation, using diffusion models as photographic instruments for what emerges in latent space. Spanning algorithmic design, model training, prompt engineering, and curatorial selection, the practice produces a herbarium of xenobotanical forms from parallel evolutionary pathways, botanical logics alien to both human and machine individually. This is not AI replacing human creativity but genuine ontological novelty arising from rigorous collaboration between silicon and carbon-based intelligence.