Introduces the first auto-regressive framework for Gaussian Splatting, enabling parallel, progressive next-scale 3D generation.
April 2, 2026
Original Paper
ARGS: Auto-Regressive Gaussian Splatting via Parallel Progressive Next-Scale Prediction
arXiv · 2604.00494
The Takeaway
It adapts the successful 2D auto-regressive generation paradigm to 3D representations, allowing for multi-scale Gaussian representations with controllable levels of detail. This significantly advances the scalability and structural consistency of 3D generative models.
From the abstract
Auto-regressive frameworks for next-scale prediction of 2D images have demonstrated strong potential for producing diverse and sophisticated content by progressively refining a coarse input. However, extending this paradigm to 3D object generation remains largely unexplored. In this paper, we introduce auto-regressive Gaussian splatting (ARGS), a framework for making next-scale predictions in parallel for generation according to levels of detail. We propose a Gaussian simplification strategy and