one dataset · paper, tables, graph, wiki
simgrammar the grammar of simulation

locator

https://arxiv.org/abs/2504.08388

notes

Visual-action autoregressive Transformer over image and action tokens with parallel decoding; 4 to 7 fps across model scales; code and model released (microsoft/MineWorld).

cited by

1 records

Diffusion and autoregressive neural game engines

provenance

runs in log/runs
2026-10-24
created2026-10-24-techniques-world-modelstechnique