LLM multimodal image generation remains too literal in the face of figurative language.
Multimodal LLMs struggle with figurative language in image generation
By
–
By
–
LLM multimodal image generation remains too literal in the face of figurative language.