ReWorld: An Interactive World Model with Long-Horizon Memory
[Sztuczna inteligencja] Autorzy: Zhifei Chen, Luozhou Wang, Guibao Shen i 8 więcej
An interactive world model must follow the user's actions, remember the places it has shown, and stream in real time. The tension is structural: control wants a short horizon, memory wants an unbounded one. ReWorld separates the two during training and bounds them at inference. Mixed per-head attention windows confine most heads to the recent past while a small set of global heads attends over the entire history, and random head routing keeps either capability from binding to particular heads; random chunk dropping makes sparse histories in-distribution. At inference the whole past lives under a fixed budget: a bounded KV cache backed by a pose-indexed landmark bank, from which the model retrieves the landmarks nearest the current pose. A metric-scale-aligned data engine places eight sources -- Unreal-rendered fly-throughs, game roaming, and real-world footage -- on one physical action …
- Kategoria arxiv
- cs.AI
- Kategoria
- Sztuczna inteligencja