TL;DR of the paper, generated automatically.

This is a survey-style report on LLM agents in computer graphics. It offers a taxonomy and a tool for mapping the literature.

• Taxonomy: The authors propose one taxonomy for agentic systems across graphics tasks, since similar ideas are often described with different terms.
• Literature system: They also built an agentic system to help researchers analyze existing and newly published papers at scale.
• Open challenges: The report covers coordinating agents and tools across tasks, interfaces for human-agent collaboration, long-horizon planning and refinement, and reliable evaluation of interactive graphics systems.
• Affiliation: The page lists the University of Texas at Austin, published Nov 1 and submitted to Hugging Face Papers Nov 3.