arXiv：cs.LG· Tianxingjian Ding, Mubarak Shah, Yu Tian·· 4 小时前AI 评分34
TERRA：通过时序效应表示与关系对齐学习可迁移的潜在动作
TERRA: Learning Transportable Latent Actions through Temporal Effect Representation and Relational Alignment
AI 导读
TERRA 通过紧凑的时序效应（净特征变化加低阶窗口内动态分量）学习连续潜在动作，并用 Effect-Anchored Transport（EAT）将潜在动作在其他初始状态下解码后锚定回源状态观测到的效应，使其由跨上下文的行为而非单一来源转移所塑造。
正文
Abstract:Latent actions supervise robot policies with action-like codes inferred from visual transitions, and their usefulness hinges on two questions: what a code keeps from a transition, and whether it still means the same thing when reused in a different initial state. The first is a tension in time: an endpoint difference discards how motion unfolds, while the full sequence admits nuisance variation. The second is left open by reconstruction, which only ever observes a latent together with the state it came from. We argue that both questions can be answered in the same place. TERRA (Temporal Effect Representation and Relational Alignment) describes a transition by a compact temporal effect, its net feature change together with a low-order within-window dynamics component, and learns a continuous latent from this effect. The same effect space then serves as the reference for reuse: Effect-Anchored Transport (EAT) decodes a latent in other initial states and anchors the resulting effect to the one observed at its source, so that the latent is shaped by what it does across contexts rather than only by the transition it came from. With frozen linear readers, TERRA predicts actions more accurately than UniVLA and a LAPA-style baseline, degrades more slowly under visual distractors, and keeps transported transitions faithful to the donor action as the recipient context moves farther away; a same-budget control shows that these gains come largely from EAT. At matched pretraining scale, the complete system reaches 93.4% average success on LIBERO, compared with 91.8% for UniVLA.
| Comments: | Preprint. Code and project page coming soon |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Robotics (cs.RO) |
| Cite as: | arXiv:2610.09509 [cs.CV] |
| (or arXiv:2610.09509v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.09509 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Tianxingjian Ding [view email]
[v1]
Wed, 7 Oct 2026 06:07:19 UTC (1,775 KB)
来源：arXiv：cs.LG · arxiv.org