SALA introduces a semantically aligned latent action framework that bridges human and robot video data without requiring paired examples. By leveraging language descriptions to estimate semantic similarities, SALA guides human latent actions toward related robot latents while grounding the latent space with robot action reconstruction. This shared latent action space enables dual‑axis scaling: larger general human pre‑training reduces robot action prediction error, and task‑specific human post‑training boosts real‑world robot success.
By Jiachun Jin, Zetong Zhou, Haonan Wang, Xiao Yang, Jun Zhu, Zhijie Deng
arXiv:2606. 00054v1 Announce Type: cross Abstract: Recent progress in generalizable embodied control has been driven by large-scale pretraining of Vision-Language-Action (VLA) models.
By Zhiyuan Feng, Qixiu Li, Huizhi Liang, Rushuai Yang, Yichao Shen, Zhiying Du, Zhaowei Zhang, Yu Deng, Li Zhao, Hao Zhao, Zongqing Lu, Oier Mees, Marc Pollefeys, Jiaolong Yang, Baining Guo
LAPT-Scale introduces a latent action pre‑training framework that leverages large‑scale egocentric human videos to train embodied robotic policies. The method comprises LAPT‑Scale‑LAM, which learns continuous latent actions by modeling visual transitions in human videos, and LAPT‑Scale‑Policy, which aligns these latent actions with robot embodiment and fine‑tunes for specific tasks. Using 23.3K hours of video, the approach improves real‑robot success rates from 63.3% to 85.0% and generalizes to unseen visual conditions.
By Zhiyuan Xiang, Peidong Liu, Mingyang Li, Jiawei Li, Wenhow Li, Wenbo Ding, Xiang Deng
World models enable agents to perform forward rollout and planning without real-world interaction. However, their application in open-world embodied intelligence remains limited by the high cost of action annotations and the heterogeneity of action spaces across platforms.
Embodied foundation models are expected to benefit from data scaling like large language models, but face a much tighter data bottleneck. Teleoperated real-robot trajectories remain the dominant pretraining source due to their precise action supervision and embodiment alignment, yet their scalability is limited by high collection cost, acquisition difficulty, and low behavioral and environmental diversity.
The paper introduces NAVA-WAM, a method that learns action priors directly from observation-only videos to pretrain action policies for robots. It uses a two-stage training process: first, it pretrains on videos with future‑video flow‑matching supervision to learn action‑relevant priors, and second, it fine‑tunes with action‑labeled demonstrations for robot control. Experiments show that NAVA‑WAM outperforms previous methods in both in‑distribution and out‑of‑distribution scenarios, achieving strong action‑label efficiency and real‑robot generalization.
By Zhaochong An, Fei Zhang, Menglin Jia, Duncan Frost, Zijian Zhou, Yikai Wang, Xudong Wang, Aditya Patel, Belinda Zeng, Tao Xiang, Serge Belongie, Amir Bar, Sen He
Zero-shot cross-task generalization, where a policy must execute manipulation tasks never seen during training, remains a central challenge in robot learning. In large language models, a novel task ca...
The paper proposes using action‑similarity supervision to improve cross‑embodiment transfer in latent action models (LAMs). By training the similarity between latent actions to match the similarity of ground‑truth robot action sequences—without predicting the actions themselves—the authors reduce sensitivity to background noise and embodiment differences. Experiments on RoboTwin 2.0 show that this approach more than doubles cross‑embodiment success compared to predicting ground‑truth actions, especially when similarities are computed on end‑effector motion and compared across robots.
By Maxime Alvarez, Renzo Caballero, Tatsuya Matsushima, Yusuke Iwasawa, Yutaka Matsuo
Zero-WAM introduces a causal video-action model that enables robots to perform unseen manipulation tasks by following in-context human video guidance. The authors create HumanGen, a dataset of 74.2K human-robot ICL pairs across 8.6K tasks, and propose an in-context future chunk prediction objective to prevent shortcut learning. In simulation, Zero-WAM attains a 47.0% success rate on seven unseen tasks, outperforming the best video-action baseline by 29.5 percentage points, and demonstrates real‑world generalization to complex, long‑horizon, and fine‑grained tasks.
By Jiaming Zhou, Qihang Zhang, Gangwei Xu, Cunxin Fan, Yujie Zhao, Ruilin Wang, Yiming Luo, Shuai Yang, Xing Zhu, Yujun Shen, Junwei Liang, Yinghao Xu
AtomEgo investigates how to integrate large-scale egocentric human interaction data into embodied foundation model pre‑training. The study uses a curated 2,659‑hour corpus and a scalable data pipeline to evaluate three co‑training paradigms across vision‑language‑action and world‑action architectures. Results show that the benefit of egocentric data depends on both its scale and the quality of alignment with robotic embodiment, offering practical guidance for scalable ego‑robot pre‑training.
By Di Wu, Dongchen Zheng, Junhe Sheng, Zhongxing Wei, Songxin Zhang, Zejian Xie, Xiaoquan Sun, Junyang Zheng, Zhuoyang Song, Jiaxing Zhang, Jiayu Chen
arXiv:2610.33685v1 Announce Type: cross
Abstract: Vision-Language-Action (VLA) models are bottlenecked by the scarcity of robot data. Egocentric human manipulation videos offer an abundant, diverse,...
By Qingyuan Cai, Bo Zhang, Yongzhen Huang, Jianyu Chen
OnlyAct proposes extracting action representations from a pretrained foundation model (V-JEPA 2) instead of learning actions from scratch. It uses a label‑free contrastive adaptation framework to isolate motion while discarding scene information, avoiding reconstruction bottlenecks. The method is evaluated on diverse benchmarks, showing more expressive action representations and better transfer of actions to new scenes.
By Runjia Li, Minghao Chen, Junyu Xie, Philip Torr, Andrea Vedaldi, Tomas Jakab