>>51445000It's literally just "magazine cover, vogue". In most models I tried this out, it reliably gens something close to VOGUE.
For other models, I've trained a LoRA. Here's its catbox (concept_vogue.safetensors):
https://litter.catbox.moe/amaxpd.safetensorsIronically, the LoRA performs worse when the model already understands the concept of vogue.