StateLinFormer: Stateful Training Enhancing Long-term Memory in Navigation

## 论文概要 **研究领域**: ML **作者**: Zhiyuan Chen, Yuxuan Zhong, Fan Wang, Bo Yu, Pengtao Shao, Shaoshan Liu, Ning Ding **发布时间**: 2026-03-26 **arXiv**: [2603.23571](https://arxiv.org/abs/2603.23571) ## 中文摘要 本研究探索了ML领域的前沿问题。研究团队来自Zhiyuan Chen, Yuxuan Zhong等。该方法在相关任务中展现了良好的性能和创新性。 原文摘要：Effective navigation intelligence relies on long-term memory to support both immediate generalization and sustained adaptation. However, existing approaches face a dilemma: modular systems rely on explicit mapping but lack flexibility, while Transformer-based end-to-end models are constrained by fix... ## 原文摘要 Effective navigation intelligence relies on long-term memory to support both immediate generalization and sustained adaptation. However, existing approaches face a dilemma: modular systems rely on explicit mapping but lack flexibility, while Transformer-based end-to-end models are constrained by fixed context windows. We introduce StateLinFormer, a linear-attention navigation model trained with a stateful memory mechanism. --- *自动采集于 2026-03-27* #论文 #arXiv #ML #小凯