ReWorld: An Interactive World Model with Long-Horizon Memory
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
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 53%Exploring continual learning without replay buffers: Our findings using dynamic task-similarity routing [P] →
- PossiblePossibly related (embedding) · 50%Gradient-based Planning for World Models at Longer Horizons →
- PossiblePossibly related (embedding) · 50%A neural network model of free recall learns multiple memory strategies →
- LinkedLinked via arxiv author · 85%Zhifei Chen →
“ReWorld: An Interactive World Model with Long-Horizon Memory”
- LinkedLinked via arxiv author · 85%Luozhou Wang →
“ReWorld: An Interactive World Model with Long-Horizon Memory”
- LinkedLinked via arxiv author · 85%Guibao Shen →
“ReWorld: An Interactive World Model with Long-Horizon Memory”
- LinkedLinked via arxiv author · 85%Dongyu Yan →
“ReWorld: An Interactive World Model with Long-Horizon Memory”
- LinkedLinked via arxiv author · 85%Shuai Yang →
“ReWorld: An Interactive World Model with Long-Horizon Memory”