This task evaluates a model's ability to depict specific architectural or engineering features. Generated images must highlight the structural elements relevant to the question rather than serve as decoration.

This task tests a model's ability to explain fundamental science phenomena (\eg, geological transformations) through instructional diagrams. Generated outputs should identify underlying mechanisms (\eg, DNA's role in genetics) or connections between concepts (\eg, how a headland erodes into caves, arches, and stacks).

This task targets a common need in academic writing, where researchers need to design figures to illustrate complex methods in research papers. The evaluated model receives a technical description of a specific method or model architecture and is asked to synthesize a self-contained figure with caption.

This task assesses whether a unified model can accurately explain complex concepts from cutting-edge computer science research in an accessible way. Given a user question about a particular method, the model must first understand the technical content and then provide a clear, coherent explanation.

This evaluates a model's ability to produce a coherent, step-by-step visual guide for drawing and creative activities.

This evaluates a model's ability to produce a coherent, step-by-step visual guide for daily life skills and household tasks.

This evaluates a model's ability to produce a coherent, step-by-step visual guide for interacting with digital devices and software systems.

This evaluates a model's ability to produce a coherent, step-by-step visual guide for physical movements and fitness routines.