HeLM: Hierarchical Explicit Language Memory for Non-Markovian Manipulation
Gives a VLA working memory by keeping task history as explicit text, updated only when events are detected. Tested on 10 real non-Markovian tasks, including ones that need context across episodes.用显式文本保存任务历史、只在检测到事件时更新，为 VLA 增加工作记忆。在 10 个真实的非马尔可夫任务上测试，其中包括需要跨回合保持上下文的任务。
BibTeX
@misc{kim2026helm,
title = {{HeLM: Hierarchical Explicit Language Memory for Non-Markovian Manipulation}},
author = {Gyeonghun Kim and Donghoon Kim and Minsoo Choi and Suyun Lee},
year = {2026},
eprint = {2611.02660},
archivePrefix = {arXiv},
primaryClass = {cs.RO},
url = {https://arxiv.org/abs/2611.02660},
}