1 paper
Andrea Morales-Garzón, Salvador López-Joya, Miguel López-Pérez +1
Vision-language models enable zero-shot classification through natural language prompts, but performance is sensitive to prompt formulation, especially in specialized domains. Zero…