VisPilot Uses Multimodal Prompting for Visualization Authoring

VisPilot addresses the imprecision of text-only instructions in automated visualization by introducing visual prompts as a complementary modality. The system reduces user iteration cycles by using visual inputs to clarify intent that natural language alone often misinterprets.