Awesome-Video-World-Models-with-AR-Diffusion

Why we included this project

This is a carefully maintained reading list for anyone tracking video world models built on autoregressive diffusion, the approach behind systems like Genie 3. Rather than a single tool you install, it organizes the field into three complementary views: algorithmic foundations, real-world applications, and infrastructure-level acceleration, which makes it a useful map for researchers and practitioners trying to get oriented. The repository updates weekly and ships a consolidated BibTeX file, so it doubles as a citation source for papers and related work, and it links to a companion framework, minWM, that walks through turning a text-to-video model into an action-conditioned world model. If you are evaluating video world modeling approaches or need a reliable map of the current ecosystem, this list saves a lot of literature-sifting time.