TY  - RPRT
TI  - Object-Centric World Models from Few-Shot Annotations for Sample-Efficient Reinforcement Learning
AU  - Weipu Zhang
AU  - Adam Jelley
AU  - Trevor McInroe
AU  - Amos Storkey
AU  - Gang Wang
PY  - 2026
UR  - https://arxiv.org/abs/2501.16443
ID  - 2501.16443
ER  -