Evaluation of vision LLMs GTP-4V and LLaVA for the Recognition of Features Characteristic of Melanoma

This dataset provides a comprehensive analysis of the ability of the multimodal large language models GTP-4V and LLaVA to recognize, classify, and appropriately comment on the ABCDE (asymmetry, border, color, diameter, and evolution) features of melanoma lesions.