arXiv cs.AI· Min Lin, Bokai Zhou, Jianqi Lin, Kaiyuan Yang, Yicheng Chang, Junjie Yu, Xiaofan Liu, Jianhua Han, Rui Li, Shen Zhao, Xiaodan Liang·· 6 天前AI 评分26
AFCE：为视觉-语言-动作模型学习未来操作目标
Beyond Actions: Learning Future Operation Targets for Vision-Language-Action Models
AI 导读
研究提出 Action–Future Coupled Effect Learning（AFCE），让 VLA 策略以未来物理 Effect 为主要预测目标，而非直接预测动作块。
正文 · 原文
Authors:Min Lin, Bokai Zhou, Jianqi Lin, Kaiyuan Yang, Yicheng Chang, Junjie Yu, Xiaofan Liu, Jianhua Han, Rui Li, Shen Zhao, Xiaodan Liang
Abstract:Actions specify what a robot should execute, but do not describe what those commands will achieve. Vision-language-action (VLA) policies nevertheless learn primarily by predicting action chunks from images, language, and state. We introduce Action--Future Coupled Effect Learning (AFCE), which makes a future-grounded physical Effect the policy's primary prediction target. An offline tokenizer constructs temporally ordered Effect tokens from demonstrated actions, realized robot responses, and sparse visual transitions. Reconstruction objectives encourage the representation to retain executable detail and observed physical change. A flow-matching VLA predicts Effects from current images, instruction, and robot state, while a state-conditioned decoder converts them into action chunks. Embodiment-specific interfaces allow single- and dual-arm tasks to share the Effect model. On DexJoCo, AFCE achieves $50.97\%$ overall success, improving over $\pi_{0.5}$ and UVT by 11.82 and 15.21 percentage points, respectively. It leads on all five bimanual tasks, highlighting the value of Effect supervision for coordinated dexterous manipulation. The project page is available at this https URL.
| Subjects: | Robotics (cs.RO); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.13212 [cs.RO] |
| (or arXiv:2610.13212v1 [cs.RO] for this version) | |
| https://doi.org/10.48550/arXiv.2610.13212 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Min Lin [view email]
[v1]
Fri, 9 Oct 2026 08:29:46 UTC (5,434 KB)
来源：arXiv cs.AI · arxiv.org