| title | WorldLines: Benchmarking and Modeling Long-Horizon Stateful Embodied Agents |
|---|---|
| authors | Yehang Zhang, Jianchong Su, Haojian Huang, Yifan Chang, Tianhao Zhou, Xinli Xu, Yingjie Xu, Yinchuan Li, Zexi Li, Ying-Cong Chen |
| date | 2026-06-17 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2606.18847 |
| upvotes | 6 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf |
| abstract | To assist humans over extended periods in real homes, embodied agents must remember user routines, world states, and past interactions. Existing long-term memory benchmarks mainly evaluate language-centric retrieval and question answering, while embodied benchmarks often focus on short-horizon task execution without testing long-term memory use in dynamic environments. We introduce WorldLines, a project-driven benchmark for long-horizon embodied household assistance. It constructs temporally extended household traces with dialogues, actions, execution feedback, object and device state changes, and converts them into evidence-linked samples for Memory QA and Embodied Task Planning. We further propose ObsMem, an observer-grounded memory framework that maintains visibility-aware memories and action-native state trails for state-aware decisions. Experiments reveal persistent challenges in partial observability, overwritten world states, and translating long-term memory into embodied plans, while ObsMem offers a stronger reference architecture for this setting. |
cite
@misc{zhang2026worldlines,
title={WorldLines: Benchmarking and Modeling Long-Horizon Stateful Embodied Agents},
author={Yehang Zhang and Jianchong Su and Haojian Huang and Yifan Chang and Tianhao Zhou and Xinli Xu and Yingjie Xu and Yinchuan Li and Zexi Li and Ying-Cong Chen},
year={2026},
eprint={2606.18847},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2606.18847}
}